Spreading the word: A social-psychological exploration of word-of

Spreading the word: A social-psychological exploration of word-of
Spreading the Word:
A Social-Psychological Exploration of Word-of-Mouth
Traveller Information in the Digital Age
Caroline Bartle
A thesis submitted in partial fulfilment of the requirements of
the University of the West of England, Bristol
for the degree of Doctor of Philosophy
Faculty of Environment and Technology
University of the West of England, Bristol
December 2011
I am indebted to the Economic and Social Research Council and the Department for Transport,
for funding this PhD. In particular, I am grateful to Lee Smith at the Department for Transport,
and Susannah Johnson at Transport Direct, for helping me to contextualise the research during
its early stages.
My thanks go to the many people who have helped me bring this thesis to fruition. Firstly, at the
University of the West of England, I wish to thank my supervisers, Erel Avineri and Kiron
Chatterjee, for their invaluable guidance throughout the PhD process; Glenn Lyons for his advice
and for helping to steer the project past its various milestones; and Derek Purdue for his support
I owe a huge debt of thanks to Toby Lewis, developer of www.bristolstreets.co.uk, who gave his
time and assistance freely, and without whom the Cycology study would not have been possible.
Thanks are also due to the many people who were interviewed for this research and who
participated in the 6-week Cycology project.
I am grateful to many members of the team in the Centre for Transport and Society at UWE for
being generous with their time, and providing help and encouragement when it was needed, but
special thanks must go to Hebba, Sendy and Charles for their academic and moral support,
particularly in the earlier stages, and to Geoff, Billy and Liz – fellow PhD students, whose good
humour and superior IT skills helped sustain me through the final furlong. One could not have
asked for a better group of colleagues.
My thanks also go to my family, especially my husband, Matt, and children, Alex and Sylvia, for
tolerating my moods and inattention at times when the PhD seemed like a demanding and
recalcitrant extra child; and finally to my father, Keith, for his careful proof-reading of the final
The use of „formal‟ travel information pertaining to costs, routes, journey times, or real-time
transport disruptions, and its role in travel behaviour (for example, choice of mode, route or
departure time) has been widely studied, but little is known about the part played by
„informal‟ information, shared through word-of-mouth amongst friends, family, colleagues and
other social networks, in relation to everyday travel. Furthermore, considerable investment
has been made over recent decades in the development of sophisticated „advanced traveller
information systems‟, delivering formal, top-down information through media such as online
journey planners, but less attention has been paid to parallel developments in the diffusion of
bottom-up, user-generated information through „electronic word-of-mouth‟ on the internet
(acknowledged in the field of marketing as a growing source of influence on consumer
behaviour). This thesis examines the role of word-of-mouth information diffusion within
everyday travel behaviour and its emerging applications in the field of online traveller
information, within a framework of social-psychological theories of behaviour and decision
theory. The exploration of social-psychological factors underlying the social transfer of
traveller information led to an expansion of existing theory, whilst the research also
generated practical recommendations for the wider incorporation of „social design features‟
into certain forms of traveller information system.
The research was undertaken in two empirical phases, both employing a qualitative
methodology. In Phase 1 (exploratory), interviews and focus groups were used to: generate
an account of the use of word-of-mouth travel information; explore participants‟ perceptions
of the influence of this form of information on their own and others‟ travel behaviour; and
identify social-psychological mechanisms underlying the influence process. „Local
knowledge‟ obtained through word-of-mouth was found to be highly valued, and was
deemed trustworthy primarily because it was based on the informant‟s direct experience (an
instrumental-reasoned explanation). However, perceived trustworthiness could be improved
by social-psychological factors such as social proximity, group-identification and accepted
norms of behaviour. Word-of-mouth was found to play a complementary role to formal
information in the decision process, and was reported to have had a direct influence on trip
details (e.g. route or departure time), but was less likely to affect modal choice. More general
interactions about travel (for example, appraising the experience of using a particular
transport mode in general conversation), whilst not necessarily perceived as travel
information per se, appeared to be influencing beliefs and attitudes, and shaping the
psychological context in which travel choices might later be made.
Phase 2 (applications) was a qualitative case-study of an innovative, web-based traveller
information system, entitled Cycology, through which 23 participants shared cycle routes and
other information with one another over a period of six weeks. This allowed both a validation
of the earlier findings within an applied context, and an exploration of some findings in
greater depth - in particular, the ways in which social norms and social identities around
travel are established or reinforced in peer-groups through word-of-mouth interactions, and
help to explain interpersonal influences on travel behaviour. Interactions on the website were
found to: influence participants‟ behaviour in the form of using cycle routes suggested by
others; strengthen pro-cycling attitudes; and enhance the experience of the cycle commute.
A key finding was the role which Cycology played in building a sense of „community‟ (group
identification), linked to high levels of trust and pro-social behaviour amongst group
members, which both reinforced positive views of cycling as a commuter mode, and
increased people‟s propensity to act on information from others within the group. Together
with the Phase 1 findings, this led to the proposed incorporation of additional „social factors‟
into established models of information use. Practical recommendations from the research
concerned ways in which developments in social media might be combined more widely
with online, map-based traveller information, particularly route-planning tools, with the
potential to enhance the perceived reliability (and influence) of such systems, and,
consequently, their effectiveness as a transport policy tool.
This copy has been supplied on the understanding that it is copyright material and that no
quotation from the thesis may be published without proper acknowledgement.
Chapter 1 : Introduction and Background ................................................................................ 1
Introduction..................................................................................................................... 1
Research Questions ....................................................................................................... 4
Background .................................................................................................................... 6
Formal and informal travel information ................................................................ 6
Word-of-mouth information diffusion .................................................................... 6
Advanced Traveller Information Systems (ATIS) ................................................. 8
UK Transport Policy Context .............................................................................. 12
Chapter Summary ........................................................................................................ 14
Chapter 2 : Literature Review .................................................................................................. 15
The role of information in the travel decision process.................................................. 15
Decision-making models in the transport literature ............................................ 16
Information-processing theory............................................................................ 17
Choice set models .............................................................................................. 19
The desire for informal types of travel information ............................................. 21
The use of word-of-mouth travel information sources........................................ 22
Social-psychological theories relevant to word-of-mouth information processes ........ 23
Social influence and the social psychology of the group ............................................. 29
ICT, social networks and travel .................................................................................... 34
The study of “online communities” ............................................................................... 36
Chapter Summary and Research Gaps ....................................................................... 38
Chapter 3 : The research approach in this thesis .................................................................. 41
Research design .......................................................................................................... 41
Qualitative research and transport: some philosophical issues ................................... 42
Philosophical foundations .................................................................................. 42
Ontology and epistemology in this thesis ........................................................... 47
Rationale for two research phases .............................................................................. 50
Chapter Summary ........................................................................................................ 50
Empirical Research, Phase 1: Exploratory Interviews and Focus Groups .......................... 51
Chapter 4 : Methodology, Phase 1 ........................................................................................... 52
Selection of methods .................................................................................................... 52
Content of the interviews and focus groups ................................................................. 56
Data analysis ................................................................................................................ 58
Cross-sectional thematic analysis ...................................................................... 59
Generalisability, reliability and validity ......................................................................... 61
Chapter Summary ........................................................................................................ 62
Chapter 5 : Findings, Phase 1 .................................................................................................. 63
The social context of word-of-mouth information-sharing: who, what and when? ....... 63
People ................................................................................................................ 63
Content ............................................................................................................... 65
Circumstances.................................................................................................... 67
The perceived role of word-of-mouth information in decision-making processes ........ 70
The role of social-psychological factors in word-of-mouth information-use ................. 73
Trust ................................................................................................................... 73
Similarity and in-group identity ........................................................................... 75
Social judgement ................................................................................................ 76
Pro-social behaviour .......................................................................................... 77
Personal norms and values ................................................................................ 77
Subjective norms ................................................................................................ 78
Social norms....................................................................................................... 79
Other areas of findings ................................................................................................. 83
Cognitive factors................................................................................................. 83
Personality-related factors ................................................................................. 84
Past (personal) experience ................................................................................ 85
Instrumental factors ............................................................................................ 86
Interest in web-based informal travel information .............................................. 87
Chapter Summary ........................................................................................................ 88
Chapter 6 : Discussion and transition to the next phase ..................................................... 90
Implications for travel information provision ....................................................... 93
Moving to the next stage: identification of areas for study in Phase 2 ............... 96
Chapter Summary ...................................................................................................... 100
Empirical Research, Phase 2. Cycology: a traveller information case-study................... 101
Chapter 7 : Methodology, Phase 2 ........................................................................................ 102
Phase 2 Research Questions .................................................................................... 102
Context: Digital word-of-mouth, online route-planning and cycling............................ 104
Methodological framework: an „experimental case-study‟ ......................................... 106
Methods of data generation ....................................................................................... 108
Selection of the case-study information system......................................................... 109
Population, sample and recruitment .......................................................................... 112
Operation of the Cycology study ................................................................................ 115
Content of questionnaires and interviews .................................................................. 116
Data analysis .............................................................................................................. 117
Establishing typologies ..................................................................................... 117
Non-cross-sectional analysis ........................................................................... 118
Chapter Summary ...................................................................................................... 119
Chapter 8 : Findings, Phase 2 - Descriptive analysis of the Cycology website ............... 120
Numerical overview .................................................................................................... 122
Content of website posts ............................................................................................ 127
Numerical analysis ........................................................................................... 127
Qualitative analysis .......................................................................................... 129
Levels of participant involvement ............................................................................... 134
Level and patterns of interaction ...................................................................... 136
Detailed analysis of two interactions .......................................................................... 141
Staple Hill to Frenchay Campus message thread ............................................ 142
Horfield to Frenchay Campus message thread ............................................... 146
Chapter Summary ...................................................................................................... 150
Chapter 9 : Findings, Phase 2 - Influence on attitudes, intentions and behaviour .......... 152
Trying out new cycle routes ....................................................................................... 152
Reinforcing pro-cycling attitudes, intentions and behaviours ..................................... 155
Taking action on cycling issues .................................................................................. 158
Changing the way people experience their commute ................................................ 160
Chapter Summary ...................................................................................................... 161
Chapter 10 : Findings, Phase 2 - The role of social-psychological factors ...................... 162
Trust in the information posted................................................................................... 164
Calculus-based trust .................................................................................... 166
Relational trust ............................................................................................. 169
Comparison with trust in other (formal) sources of information ................... 174
Informational, normative and referent social influence ................................ 175
Group identification („community‟) .............................................................................. 175
A community of cyclists ................................................................................ 176
A workplace community ............................................................................... 179
A project community ..................................................................................... 180
Mutual support (solidarity) ............................................................................ 181
Shared experiences and emotions (empathy) ............................................. 182
Lack of group identification .......................................................................... 183
Effects of limited group size and composition ............................................................ 184
Positive effects ............................................................................................. 185
Negative effects............................................................................................ 187
Social judgement ........................................................................................................ 188
Judgements about the group ....................................................................... 189
Judgements about individuals ...................................................................... 190
Pro-social behaviour and attitudes ............................................................................. 192
Reasons for contributing .............................................................................. 192
Reasons for not contributing ........................................................................ 197
The role of other („non-social‟) factors ....................................................................... 200
Instrumental factors ...................................................................................... 200
Individual factors: personality and past experience ..................................... 202
Reflections on the Phase 2 methodology .................................................................. 204
Chapter Summary ...................................................................................................... 208
Chapter 11 : Implications for the wider field of advanced traveller information .............. 210
Generalising from the case-study .............................................................................. 211
Desirable „social design features‟ .............................................................................. 212
User recommendations ................................................................................ 213
A system developer‟s perspective ................................................................ 218
Potential relevance to other transport modes ............................................................ 220
User groups: desirable size and composition ............................................................ 225
Developing a user typology .......................................................................... 229
Four individual cases (of user) ..................................................................... 232
Chapter Summary ...................................................................................................... 238
Chapter 12 : Conclusions ....................................................................................................... 239
Theory development .................................................................................................. 239
Advanced traveller information systems .................................................................... 243
Discussion: relevance to behavioural change policies and measures....................... 248
References ................................................................................................................................ 251
Appendix A – Phase 1 Interview Topic Guide ....................................................................... 267
Appendix B – Phase 1 Focus Group Topic Guide ................................................................ 273
Appendix C – Exploring social-psychological factors: examples from research ............. 278
Appendix D – The influence of theory-driven analysis methods ........................................ 283
Appendix E – List of NVivo codes, Phase 1 .......................................................................... 286
Appendix F – Post-Cycology Questionnaire ......................................................................... 290
Appendix G – Cycology interview guide template ............................................................... 292
Appendix H – Cycology Guide for Participants .................................................................... 296
Appendix I – Information sheet for Cycology participants .................................................. 302
Appendix J – Code of conduct for on-line discussions ....................................................... 304
Appendix K – Categories of website post over time ............................................................ 306
Appendix L – List of NVivo codes, Phase 2 .......................................................................... 308
Appendix M – Titles of posts and threads added to the Cycology website ....................... 312
Figure 1-1: Overview of PhD Structure .................................................................................... 5
Figure 1-2: Formal and informal travel information .................................................................. 8
Figure 1-3 Bristolstreets website (screenshot) ....................................................................... 11
Figure 2-1: Model of the travel choice destination process, adapted from Um and Crompton,
1990. ...................................................................................................................................... 20
Figure 2-2: Triandis‟ Theory of Interpersonal Behaviour (source: Jackson, 2005) ................ 25
Figure 2-3: Contribution of social psychology theories and constructs to understanding of the
topic ........................................................................................................................................ 33
Figure 5-1: People - giving and receiving information before/after a trip ............................... 64
Figure 5-2: People - giving and receiving travel information during a trip ............................. 65
Figure 5-3: Social-psychological factors: motivations for giving information to others .......... 81
Figure 5-4: Social-psychological factors affecting use of information received from others .. 82
Figure 6-1: The role of word-of-mouth information in the pre-trip travel decision process .... 95
Figure 7-1: Screenshot showing “About” section of www.bristolstreets.co.uk ..................... 111
Figure 8-1: Screenshot of the Cycology website at the end of week 1. ............................... 123
Figure 8-2: Number of messages posted over time ............................................................. 125
Figure 8-3: Number of markers viewed over time ................................................................ 125
Figure 8-4: Number of posts by type .................................................................................... 128
Figure 8-5: All route-related comments compared with all other comments over time........ 129
Figure 8-6: Writing posts and opening markers, by individual ............................................. 135
Figure 8-7: Number of posts in „threads‟ and number of stand-alone posts written by each
participant ............................................................................................................................. 137
Figure 8-8: Posts written per person compared with number of times their posts were read by
others ................................................................................................................................... 138
Figure 8-9: Who opened whose markers – whole project ................................................... 139
Figure 8-10: Who opened whose markers – North West cluster ......................................... 140
Figure 8-11: Who opened whose posts: South West cluster ............................................... 141
Figure 8-12: Viewings of the "Staple Hill to Frenchay Campus" thread............................... 143
Figure 8-13: Screenshot: Staple Hill to Frenchay Campus .................................................. 144
Figure 8-14: Viewings of the "Horfield to Frenchay Campus" thread .................................. 147
Figure 8-15: Screenshot: Horfield to Frenchay Campus route ............................................ 148
Figure 9-1: Effect of shared information on participants' route choice ................................. 153
Figure 10-1: Interlocking social-psychological themes ........................................................ 164
Figure 10-2: Factors contributing to trust in the information posted on Cycology ............... 166
Figure 10-3: Interlocking concepts of „community‟ in Cycology ........................................... 177
Figure 10-4 Hierarchy of pro-social behaviour in Cycology ................................................ 193
Figure 11-1: Types of Cycology participant ......................................................................... 230
Figure 12-1 : Processing of informal and formal travel information ..................................... 241
Figure 12-2: Mock up of a cycle route planner incorporating user comments ..................... 246
Table 4-1: Composition of Phase 1 sample by age, gender and primary transport mode for
commuting to the university ................................................................................................... 55
Table 7-1: Cycology participants by age, gender and organisation ..................................... 114
Table 7-2: Cycology participants – commencement and frequency of cycling to work ....... 114
Table 10-1: Stated reasons for low level of posts (seven lowest contributors) .................... 199
Table 11-1: Recommended „social design features‟ ............................................................ 219
Table 11-2: Potential use of online information-sharing for other (non cycling) transport
modes................................................................................................................................... 224
Introduction and Background
Chapter 1 : Introduction and Background
This chapter introduces the subject area of the research, presents the objectives and research
questions, and a schematic overview of the structure of the PhD. The concepts of „informal‟ and
„formal‟ travel information and word-of-mouth diffusion – as they are to be used in this thesis are then defined, and background provided on contemporary technical and policy developments
in the field of traveller information.
Today‟s traveller has a wealth of information at his or her disposal to help plan and execute dayto-day journeys, from old-fashioned paper maps and timetables, to state-of-the-art satellite
navigation systems and web-based services delivering detailed journey planning and real-time
travel updates to computer or mobile phone. Conventional rational choice explanations of human
behaviour would contend that the modern traveller, armed with comprehensive information about
all her travel options, as well as powers of perception and reasoning, weighs up the costs and
benefits (utilities) to herself of each option and arrives at a „rational‟ decision to act in a way
which she expects will bring her the maximum personal benefit. Thus, access to detailed, factual
travel information should, through a cognitive process of „preference maximisation‟, lead
individuals to make optimal travel decisions (e.g. the cheapest/quickest/least congested route or
mode), which may, when aggregated, lead to improvements in the efficiency of the transport
system as a whole. Furthermore, integrated, multi-modal travel information facilitates the
planning and execution of journeys by public transport, and it is suggested that this could
encourage some people to make a journey by public transport, when they would otherwise have
used the car, by raising their awareness of alternatives and reducing the time and effort of
information search. This has clear policy advantages in terms of reducing traffic congestion and
its negative social and environmental consequences.
Although deeply embedded in the culture and institutions of Western society, the behavioural
assumptions inherent in the rational choice model have been the target of much criticism in
recent decades, particularly from within the fields of psychology and sociology. Criticisms have
been levelled against three key assumptions:
1) that choice is rational, in the sense of a continual deliberation to maximise utility;
2) that the individual is the appropriate unit of analysis in social action;
3) that choices are made in the pursuit of individual self-interest.
Introduction and Background
A strict interpretation of the first assumption may lead one to overlook the significance of factors
such as emotion, habit and an individual‟s cognitive limitations; the second neglects the effect of
social factors on human agency, such as social norms, social expectations and group identities;
and the third may fail to acknowledge the influence of personal (moral) norms and pro-social
values (Jackson, 2005).
In the context of everyday travel choices, therefore, it may be naïve to expect the provision of
more and better information to lead directly and inevitably to changes in travel behaviour which
benefit the individual, as well as society as a whole (such as reductions in traffic congestion).
Indeed, the evidence of actual behavioural change as a direct result of the use of travel
information is more limited than one might expect, particularly in terms of modal shift. Take, for
example, the case of Transport Direct (the UK multi-modal travel information service). Despite
an early ambition to “enable and encourage people to use public transport more than they
currently do” (DfT, 2002) through the provision of multi-modal information demonstrating
alternatives to the car, an online survey showed that 7.7% of users, after consulting the
information service, intended to switch their mode from car to public transport for a trip they had
made before, and 2.3% intended to switch the other way. This suggests an intended net modal
shift to public transport of just 5.4% (AEAT, 2007).Other studies have suggested that presenting
people with factual information is insufficient on its own to motivate behavioural change (e.g.
Ampt, 2003). In the context of engaging the public in environmental policies, Owens (2000)
argues that:
“A substantial body of social-scientific research suggests that, while greater knowledge
may be worthwhile in its own right, barriers to action do not lie primarily in a lack of
information or understanding (…). We should not be greatly surprised to find (. ..) that
cultural rules and social networks had a greater influence on translating underlying
concern into environmental action than the availability of detailed scientific information”..
This may lead us to wonder whether the same argument might apply in the field of travel
information and behaviour. Does informal information, shared through word-of-mouth amongst
friends, family, colleagues and other social networks, have a role to play in travel decision
making, and if so, might this be explained with the help of behavioural theories and constructs
from the social sciences (particularly social psychology)? Although social influence might be
most potent within immediate peer groups, experimental social psychology has demonstrated
that individuals are also influenced by people they do not know personally, but with whom they
feel some sense of connection, however fleeting. This might be, for example, an affinity with
fellow cyclists, users of the same bus service, or even a brief sense of identification with fellow
Introduction and Background
passengers on a station platform when delays are announced. A number of social-psychological
theories might be drawn upon to identify and explain some of the social factors underlying wordof-mouth information-use and its potential effects on behaviour, using concepts such as social
learning (Bandura, 1977), social norms (Triandis, 1977; Cialdini et al., 1990) social identity
(Tajifel, 1982) and referent informational influence (Turner et al., 1987).
A further dimension to this area is the growth of “digital word-of-mouth” (Dellacoras, 2003). The
ability to connect with „people like me‟, unimpeded by geographical distance, has radically
increased as a result of the explosion in electronic communications, particularly internet-based
social networking. This has rapidly expanded the realm of word-of-mouth information diffusion
from traditional face-to-face communication, greatly increasing the channels for exchanging
„bottom-up‟ information. In fact, many user-generated travel and transport-related web-sites and
email groups already exist for the purpose of sharing opinions and recommendations amongst
groups of like-minded travellers. These are developing in parallel with increasingly sophisticated
advanced traveller information systems. However, so far there has been little interaction
between these two forms of information delivery. Web-based journey planners do not currently
link to the type of customer reviews accessible to people buying, for example, books or holidays
over the internet, which could address the informal travel information needs which might
sometimes be helpful in planning a journey, although some public transport companies have
recently started to embrace social networking, and there have been a number of developments
in web-based interactive mapping. If the underlying aim of advanced traveller information is to
help people make better-informed decisions which benefit not only themselves, but also the
efficiency of transport networks as a whole (which may imply greater use of public transport to
mitigate traffic congestion), then the inclusion of user-generated, informal travel advice might
prove to be one way of widening its impact.
The aims of the research were thus defined as follows:
To explore the role of informal information and word-of-mouth diffusion in everyday
travel behaviour;
To identify and explore social-psychological factors which may account for word-ofmouth influences on travel behaviour;
To study the above phenomena in the context of an innovative, interactive information
system, and to consider implications of word-of-mouth information-sharing for a wider
range of advanced traveller information systems.
Introduction and Background
In summary, this research has offered an opportunity to combine theoretical concepts from a
number of disciplines with empirical research to improve understanding of an unexplored
phenomenon in transport studies, and to apply this knowledge to a specific area of transport
policy and practice. The research takes place against a backdrop of rapid technological change
in the field of advanced traveller information and user-generated content. The findings might
prove to be of interest to developers and sponsors of such systems, and to policy-makers
concerned with maximising the effectiveness of advanced traveller information in the context of
broader measures to tackle traffic congestion.
Research Questions
The research questions were developed from the original research aims and refined iteratively
as the literature review progressed.
a) Context: the social transfer of travel information.
1. From whom, and in what circumstances, do people acquire travel information, and with
whom do they share it? What types of information are conveyed through word-ofmouth?
b) Mechanisms: the influence of word-of-mouth information in travel behaviour.
2. What is the perceived role of word-of-mouth information in the travel decision process?
How do informal and formal types of information interact?
3. How do particular social-psychological factors (e.g. social norms, social identity, prosocial values, trust) appear to influence the use and effects of word-of-mouth
c) Applications.
4. How do word-of-mouth processes and informal information-sharing operate within an
emerging web-based system incorporating user-generated travel information?
5. How might information-sharing amongst travellers be applicable to advanced traveller
information systems (ATIS) more generally?
The empirical research in this thesis was undertaken in two phases. Phase 1 (exploratory)
addresses the social context of travel information (research question 1) and underlying socialpsychological mechanisms which might help to explain the role of word-of-mouth in travel
decision-making (research questions 2 and 3). In Phase 2 (applied), the initial findings were
Introduction and Background
translated into and explored more deeply within a specific context – a case-study of a webbased system enabling users to share travel information with one another (research question 4).
Finally, the thesis broadened out once more by developing preliminary recommendations, based
on the findings from both research phases, for the incorporation of „social design features‟ into
advanced traveller information systems more generally (research question 5).
A „road map‟ of the PhD structure is provided in Figure 1-1, showing the two phases of the
research and their relationship with the research questions (RQs) and research methods.
Figure 1-1: Overview of PhD Structure
Literature Review
Information and travel
Behavioural theory
(social psychology)
diffusion and ICT
Phase 1: Exploratory research
RQ 1
RQ 2 & RQ 3
Phase 2: Applied research
Interviews and
focus groups
ICT Case-study
RQ 4
Implications for
RQ 5
Introduction and Background
This part of the chapter begins by defining both formal and informal travel information, and wordof-mouth information diffusion, as they will be used in this thesis. It then outlines two aspects of
the research context: the state-of-the-art in advanced traveller information systems and usergenerated content at the time of writing; and the UK policy context in relation to traveller
information provision.
Formal and informal travel information
This study differentiates between formal and informal travel information, with regard to both
information content and the modes through which information is transmitted. In terms of content,
„formal travel information‟ may be described as „factual‟ information provided to a mass audience
(although it may be tailored to the needs of the individual) by, for example, transport providers
and governmental agencies. It comprises information such as: departure and arrival times;
costs; possible routes; and real-time updates on the running of the transport network. Formal
modes of information delivery may be described as „top-down‟ communication mechanisms and
may include: paper timetables and route maps; telephone inquiry services; teletext and digital
TV; GPS navigation systems; and web-sites delivering journey planning and real-time travel
By contrast „informal travel information‟ comprises what might be described as „softer‟ aspects of
the travel experience. It may cover many topics, from the perceived reliability of transport
services, ease of interchange, perceptions of personal safety, ease of taking luggage, likelihood
of getting a seat, to scenic aspects of the route. It is therefore rather subjective information likely
to arise from the personal experience and attitudes of the information-giver, and orientated to the
particular interests and needs of the information-recipient. It is likely to be regarded by the
participants as „knowledge‟ rather than information (cf. Kenyon and Lyons, 2003), and may be
combined with an attempt to influence other peoples‟ attitudes, values and behaviour. Informal
modes of information delivery are encapsulated by the term „word-of-mouth‟.
Word-of-mouth information diffusion
Informal delivery of information may be described as the passing of information, especially
recommendations, through word-of-mouth in an informal person-to-person manner. A concept
commonly used in marketing research and practice, word-of-mouth is traditionally defined as
one-to-one and face-to-face exchange of information about a product or service, although this
definition has been widened to embrace one-to-one interactions conducted at a distance via
Introduction and Background
communications technologies (Godes et al., 2005; Dellarocas, 2003). Moreover, information
exchanged in online communities shares many common features with traditional word-of-mouth,
although it is not strictly one-to-one, but one-to-many in nature. Hence, it can be argued that
informal, or word-of-mouth, diffusion processes are in operation when travel advice is shared via
social networking websites, just as they are when people interact with one another face-to-face.
Thus, we find that formal travel information may be transferred either through „top-down‟
mechanisms (via published timetables etc.) or through word-of-mouth. Informal travel
information, however, is usually transferred through word-of-mouth, although advanced traveller
information systems would seem to offer the possibility of combining both types of information.
The different combinations are shown in Figure 1-2. Previous research on the use and effects of
travel information has concentrated on factual information, delivered through formal („top-down‟)
means such as published maps and timetables (upper left quadrant in Figure 1-2). Although
research on travel information awareness and use has touched on word-of-mouth sources such
as “asked a friend” (e.g. UK National Travel Survey, 2005 and 2006), the social transfer of
formal travel information has not been explored in detail. Research in the field of tourism and
leisure travel, and consumer studies, has shown the significance of word-of-mouth, in the travel
decision-making process (e.g. Gretzel et al., 2007, Murphy et al., 2007, Bieger and Laesser et
al., 2004), but similar research has not so far been undertaken with reference to everyday
journeys. Moreover, the tourism literature has been more concerned with decisions about
destinations than with the process of travelling to and from these destinations. This thesis is
structured in such a way that the two right hand quadrants in Figure 1-2 will constitute the main
scope of research in Phase 1. In Phase 2, the areas in the right-hand quadrants are explored in
the context of an online information system where both formal and informal types of information
are shared through word-of-mouth. Finally, the study moves into the domain of the lower lefthand quadrant by considering the incorporation of informal (user-generated) information into
„formal‟ advanced traveller information systems.
Introduction and Background
Figure 1-2: Formal and informal travel information
Mode of Information Delivery
Content of Information
Factual travel information
delivered through top-down
Factual travel information
delivered through wordof-mouth
Informal travel advice
delivered through top-down
Informal travel advice
delivered through word-ofmouth.
Advanced Traveller Information Systems (ATIS)
Modern advanced traveller information systems are extremely varied in form and content,
offering journey-planning, real-time travel information and navigational advice through a range
of technological media. Common examples include: detailed web-based information services; incar navigation systems; and electronic message boards positioned beside motorways, bus stops
and train platforms. Most systems of this type aim to provide the traveller with factual information
such as journey times, costs, real-time arrivals and departures, and warnings of disruption on
transport networks. Those systems to be given the greatest attention in this study are webbased journey-planning and real-time information systems, accessible via computers, mobile
phones and other portable devices, as this is a technological medium which offers considerable
scope for social interaction amongst users, as well as potential for the inclusion of usergenerated content.
Introduction and Background
The opportunities for users to share travel information and advice via these systems are
currently rather limited. For example, some web-based journey planners, such as those offered
by Transport for London, Transport Direct and the AA, offer the user the opportunity to “email
this route to a friend”, but the assumption appears to be that users are generally seeking travel
information for their own individual use, without the desire to interact with other users. According
to the categorisations in Figure 1-2, these systems could therefore be said to provide formal
information via formal delivery mechanisms.
One example of a journey planning service which includes an element of informal advice is
www.seat61.com, describing itself as an “independent, personal website” which provides links to
public transport information around the world. This website features some of the „softer‟
information about journeys (including photographs) which are missing from other journeyplanners, such as ease of dealing with particular interchanges, and matters of comfort on
particular transport services. This informal advice is based on the personal experience of both
the website owner - “the man in seat 61” and contributions from website users. However,
contributions are managed by the website owner, rather than being user-generated. The focus of
the website is international leisure travel. Existing user-generated websites, also known as Web
2.0 or peer-to-peer (P2P) sites, relating to travel, such as www.tripadvisor.com, tend to focus on
destinations rather than journeys, and again mainly address leisure travel. At the same time,
some travel websites have converged with social networking, such as www.dopplr.com which
helps long-distance business travellers to share their travel schedules. Another (leisure) travel
website has incorporated information provided by local residents: www.spottedbylocals.com, but
like www.seat61.com, this information is managed by the website owner.
Discussion around day-to-day travel is found more commonly via an abundance of specialist
online forums (e.g. discussion forums hosted by local cycling campaigns such as
www.bristolcyclingcampaign.org.uk or national motorists‟ forums such as the RAC‟s
www.rac.co.uk/web/forum). Social networking sites such as Facebook enable members to join
groups linking them to others with a shared interest (e.g. the Facebook “Friends of Bristol
Streets” group). A further example is the cyclists‟ social network: http://morvelo.cc/. Another
recent development at the time of writing is location-based social networking applications for
mobile phones, such as Foursquare, which allows people to leave location-related tips (for
example, a recommended café) for friends in their network who may be in the same area. At the
time of writing a further innovative project is underway in Brighton; in the TWAGO project
(“Twitter As you Go”), local residents are using the micro-blogging system Twitter to share their
local travel experiences. Recently, some public transport companies (particularly buses) have
Introduction and Background
begun to set up Facebook pages both to inform customers and elicit feedback; this development
will be returned to in Chapter 11.
One area where some convergence between formal and informal information content has been
identified is that of interactive mapping. This is where web-based maps, usually based on
Google maps, but also bottom-up initiatives such as www.OpenStreetMap.org, are overlaid with
route and other travel information, such as location of bus stops and timetable information
(formal information). However, informal information can also be added by users, who may, for
example, draw on their preferred cycle routes and post up comments and photographs. One
example of a system of this type is www.bristolstreets.co.uk (Figure 1-3). Although not strictly a
traveller information system (it also offers many other types of local information), this website
offers an interesting example of how formal and informal types of travel information might be
combined. A cycling layer of the site invites users to draw routes onto the map to share with
others, as well as add comments about cycling facilities. This website was eventually selected
as the platform for the Phase 2 case-study, and will be returned to in greater detail in Section
www.OpenStreetMap.org is another P2P mapping project which allows users to create and
share geographic data such as street maps, as well as to undertake route planning and add
comments via a web log. Examples of informal travel information-sharing within online
communities are the cycling journey planner developed by the Cambridge Cycling Campaign at
www.camcycle.org.uk, which allows users to upload photographs to the map; this is now offered
throughout the UK in the form of www.cyclestreets.net. Further examples are the interactive map
developed by the Camden Cycling Campaign: www.camdencyclists.org.uk and Transport for
London‟s cycle web pages at http://www.tfl.gov.uk/ which both now incorporate user-generated
cycle routes. A more detailed discussion of current developments in online route-planning for
cyclists is provided in Section 7.2 as context for the Phase 2 research.
This section has provided an overview of some of the „social” elements of existing web-based
sources of traveller information, which provided a context for the case-study undertaken for
Phase 2 of the thesis, and which will be returned to in Chapter 11, when potential applications of
the research will be considered. The remainder of this chapter will now turn to the UK policy
context of advanced traveller information.
Introduction and Background
Figure 1-3 Bristolstreets website (screenshot)
Introduction and Background
UK Transport Policy Context
Providing better traveller information is one of the themes set out in the Department for
Transport‟s policy framework for Intelligent Transport Systems (DfT, 2005), in which the
following policy goal is articulated:
“facilitate the provision of and provide accurate, timely and relevant information to
travellers so that they can make informed choices about travel mode and time and as
part of the contribution to tackling congestion and enabling more efficient and less
stressful journeys” (DfT, 2004, p.28)
More specifically, UK government interest in advanced traveller information is perhaps best
exemplified by its investment in Transport Direct, the integrated, multi-modal travel information
service announced in the 10 Year Plan for Transport (DfT, 2000), and launched in July 2004. By
mid-2007, £55 million had been spent on the development of the service. It emerged from a
policy discourse around the desirability of creating “seamless journeys” by public transport,
whereby public transport might be presented as an attractive alternative to the convenience of a
car. The 1998 Transport White Paper (DfT, 1998) identified better public transport information as
one of the tools which might help to achieve the government‟s objective of greater public
transport use. Consistent with this policy approach, the Outline Transport Direct Business Plan
(DfT, 2002) stated that “Transport Direct will help boost the public transport industry and enable
and encourage people to use public transport more than they currently do” (p.4). The role of
public transport information as a “soft” transport policy measure designed to encourage
behavioural change was one of the measures explored in a study carried out for the Department
for Transport (Cairns et al., 2004) and referenced in the 2004 Transport White Paper (DfT, 2004,
p.40). “Soft” transport policy measures, also termed “smarter choices”, seek
to encourage people to use their cars less by enhancing the attractiveness of transport
alternatives, including through better information provision. At the time of its inception, therefore,
Transport Direct was presented as one of a package of policy tools designed to ease congestion
on transport networks by stimulating behavioural change. Later, however, the overt message
about the purpose of Transport Direct moved away from “getting people out of cars”, towards
offering people greater transport choice . Stakeholders have expressed the view that lack of
information is not the main barrier to modal shift, so this service should not be expected to lead
DfT reply to parliamentary question from Rose Kramer MP, 8 October 2007.
Conversation with Transport Direct representative, August 2007.
Introduction and Background
directly to this form of behavioural change (AEA Technology, 2007). However, at the same time,
Transport Direct continues to expand its provision of information about alternatives to the private
car, exemplified by the recent addition of a cycling journey planner, which at the time of writing
covers 32 towns, cities or areas; approximately £2.4 million had been spent on or allocated to
this feature by the end of the 2010/11 financial year .
The 2011 Local Transport White Paper, Creating Growth, Cutting Carbon (DfT, 2011) continues
the discourse of enabling choice through information provision:
“The concept of enabling choice following provision of better information and education
is what the Government‟s approach to sustainable travel is all about”. (p.13)
However, the 2011 White Paper also incorporates the concept of “nudge” into information
provision as a tool for behavioural change. “Nudge” (Thaler and Sunstein, 2009) involves
encouraging individuals to make „good‟ decisions by presenting choices in a way which nudges
them in a particular direction. Nudging people towards more sustainable travel behaviour may
involve (among many other measures) the provision of information about sustainable travel
choices which is not only comprehensive, but also framed in a manner designed to encourage
particular choices – for example, through careful use of language. This might be interpreted as a
return to the earlier conceptualisation of travel information as a tool for promoting not just choice,
but the „right‟ choice (c.f. the 1998 Transport White Paper), albeit in a more subtle manner.
The Transport Direct cycle journey planner is also promoted in the 2011 White Paper as one
means of encouraging “active travel”: “In particular we aim to remove one of the key barriers for
novice cyclists – knowing where it is easy and safe to ride” (p.48). The use of the cycle journey
planner is particularly encouraged within local contexts such as: schools; social groups; railway
stations; hospitals and universities. Moreover, local authorities are encouraged to promote it as a
workplace travel planning tool. This provides a highly relevant backdrop for the case-study
research carried out for Phase 2 of this thesis (featuring an online information system for
cyclists); policy relevance of the research will be returned to in the concluding chapter of the
DfT Response to Freedom of Information Act request for expenditure on the “find a cycle route
feature” of Transport Direct. FOI Reference number F0006273.
Introduction and Background
Chapter Summary
This chapter has introduced the thesis by setting out the intellectual context of the research: how
the conventional paradigm of information-use by the individual, rational decision-maker in
pursuit of utility maximisation might be called into question with the benefit of insights from the
social sciences, particularly social-psychological theories of behaviour. What happens when
information incorporates not just „facts‟, but subjective experiences and personal advice as it is
communicated between people? Research suggests that information provision on its own has a
limited effect on behaviour, but might information obtained through word-of-mouth include an
element of social influence which might increase its impact? If so, might informal, word-of-mouth
information have a role to play in travel information provision – particularly in the domain of
internet-based traveller information systems, where opportunities for the incorporation of usergenerated content continue to grow? This chapter has presented the objectives and research
questions, defined some of the key concepts within the thesis, and set them within the context of
current technological and policy developments in the field of traveller information.
Literature Review
Chapter 2 : Literature Review
This chapter summarises relevant aspects of five areas of literature which helped to locate the
research to be carried out for this thesis, identify specific research gaps, and inform the
development of a conceptual framework by identifying relevant theories and constructs. Firstly,
Section 2.1 presents a summary of the literature on the role of information in the travel decision
process, which typically focuses on „formal‟ information and regards information-use as an
individual matter. The review then turns to those studies which have identified a wish among
travellers for informal types of information, and those which have made reference to word-ofmouth information diffusion (substantive work in this area having been carried out in relation to
tourism and leisure travel, but not utility travel). Having identified a gap in knowledge about
social aspects of information-use within the transport literature and within information-processing
models, the review then turns in Section 2.2 to areas of social-psychological theory which have
been applied to travel choice behaviour - particularly those which emphasise the importance of
social interactions, in order to identify theories which might help to explain social aspects of
information-use. Section 2.3 explores one area of social-psychological theory in greater detail:
social influence within the group (important for the interpretation of the case-study findings).
Section 2.4 considers literature from the disciplines of sociology and geography, exploring new
relationships between communications technologies, social networks and travel, and
implications this might have for the word-of-mouth diffusion of travel information through social
networks. Finally, Section 2.5 summarises some recent research into online communities, which
was informative in the development and interpretation of the Phase 2 case-study.
The role of information in the travel decision process
The research literature relating to the use and effects of travel information provided an essential
starting point for this study (cf. Lyons et al., 2007, in which over 100 articles published since
2001 were reviewed). Of particular interest were those studies relating to the role of information
in travel decision-making, the different types of travel information which people need or would
like, and the extent to which people seek information through word-of-mouth. Most of the
literature, however, focuses on individual decision-making, without considering in detail the role
of social interaction in the way people acquire and use travel information. Among the research
gaps identified by the Lyons et al. review (2007) were: better understanding of the social context
of information-use, and an exploration of ways in which social factors might be better
incorporated into information services. The current research was construed partly as a means of
helping to fill these research gaps.
Literature Review
Decision-making models in the transport literature
A consideration of some of the decision-making models which have been applied within the
literature to explore the role of information in travel decision-making provided a first step in the
development of a theoretical framework for this project. Four commonly-used approaches in this
field are: utility maximisation; bounded rationality; “satisficing behaviour” (decision-making which
leads to satisfactory rather than optimal results); and habit. These approaches were reviewed by
Chorus et al. (2006a). These are essentially models of individual behaviour which do not
emphasise social influences on decision-making. However, they provide a number of conceptual
approaches within which social factors might be considered.
The utility maximiser in micro-economic theory rationally weighs up the costs and benefits of all
options in a continual process of deliberation, and makes the choice which best serves his or her
individual interest. Access to complete information is assumed. This view of human decisionmaking is increasingly questioned, not least because it does not incorporate social factors into
the concept of utility, and has been found to be a poor predictor of actual travel behaviour
(Gärling, 1998). The concept of bounded rationality (Simon, 1959) acknowledges that whilst
people may aim to maximise their utility, they also have cognitive limitations which may restrict
their ability to understand or remember information. They may therefore employ “satisficing
behaviour”: seeking sufficient information to reach a travel decision which is good enough to
satisfy a certain level of expectation. They may also take short-cuts, or heuristic decisions, and
this limits their motivation to seek or act upon travel information (Tversky and Kahneman, 1974).
The role of habit in travel choice has been widely studied (e.g. Verplanken et al., 1997, Bamberg
et al. 2003); habitual behaviour, defined as behaviour without deliberation, usually precludes the
seeking of travel information except for confirmatory purposes, although habitual travellers may
be forced to seek information when their usual travel arrangements are disrupted. Although
social factors are not explicitly addressed within these approaches, they may prove to be
significant; for example, social norms and social learning might encourage heuristic decisions, or
reinforce habitual behaviours.
It is important to sound a note of caution in treating utility maximisation in decision-making as
distinct from „less rational‟ models. For example, in some circumstances it might be „rational‟ to
make a heuristic decision, because the costs of seeking further information outweigh the
benefits. Moreover, a decision which seems irrational in isolation might actually contribute to
utility maximisation at a longer-term strategic level. It is clear that long-term decisions concerning
lifestyle and mobility, such as residential or work location, or car-ownership, place constraints on
(or indeed create opportunities for) short-term daily activity planning decisions, which in turn
Literature Review
affect the very short-term rescheduling decisions which may arise when unforeseen events
occur (Ben-Akiva et al., 1994, cited in van der Horst, 2006).
Market research by Social Research Associates (SRA) (2004 and 2005) applied some of the
theoretical literature on decision-making to the case of Transport Direct. This work identified a
broad range of decision-making models, and their relationship to information-use, as well as
assessing implications for the design of Transport Direct itself. For example, the assumption that
decision-makers are purely rational would imply the provision of a mass of information, from
which users would select the optimum, utility-maximising choice. The models identified in this
report include: „problem-solving‟, which corresponds with the satisficing model described above;
the “incremental model”, in which habit plays a strong role; and “discrete choice”, which the
authors link to bounded rationality. Of particular interest for this research is the “steering” model,
which suggests that some people rely heavily on recommendations from others. Interestingly,
the same people often employed different strategies depending on the circumstances. This
finding is supported by consumer research demonstrating that the same individual may use a
variety of different decision strategies (Bettman et al., 1998); it suggested that care should be
taken in this thesis not to categorise research participants as particular types of decision-maker
relying on one or other type of information in all situations.
Information-processing theory
There is a long history of research into the role of heuristic rules in information-seeking
behaviour within disciplines such as consumer psychology and behavioural economics (e.g.
Stigler, 1961), which has been particularly influential in the development of consumer choice
models. One of the “grand models” of consumer decision-making is information-processing
theory (Bettman, 1979, cited in Gabbott and Hogg, 1994; Jackson and Polak, 1997; Sirakaya
and Woodside, 2004; van der Horst, 2006). This was developed some thirty years ago as an
alternative approach to rational choice theory, incorporating concepts such as bounded
rationality (Simon, 1959). More recently, this perspective has been extended to provide an
adaptive decision-making framework: individuals adapt their decision-making strategies in
response to situational constraints and the complexity of the decision task (Payne et al, 1993).
Consumer behaviour theorists have argued that consumers frequently construct rather than
simply reveal their preferences when faced with choices of any novelty or complexity (Bettman
et al. 1998).
According to information-processing theory, the consumer decision-making process comprises
five main stages: (1) problem recognition; (2) information search; (3) alternative evaluation and
selection; (4) implementation; and (5) post-purchase evaluation (van der Horst, 2006; Sirakaya
Literature Review
and Woodside, 2005). Bettman (1979) suggests that during this process the consumer is
influenced by the following psychological, or individual, factors: motivation, goals and internal
information (experience). Other theorists have extended the range of individual factors to
include, for example, beliefs and attitudes, which may exert an unconscious influence on the
decision-making process (Sirakaya and Woodside, 2005).
The stage most pertinent to the present study is information search, which psychologists divide
into two categories: “internal” and “external” search. Faced with a decision “problem” (i.e. a goal
which they wish to pursue) individuals initially carry out an internal search, drawing on past
personal experience and memory; only if this process yields insufficient information will they go
on to execute a search of external information sources. In the context of this study, word-ofmouth is one of the ways in which people may seek external information. Internal information
takes precedence over external information because personal experience is rated more highly
by the individual (Avineri and Prashker, 2006), partly because stored information is believed to
be more reliable, and partly because this process does not incur additional financial or emotional
costs. Jackson and Polak (1997) point out that the preference for stored information goes some
way to explain habitual travel behaviours. In the context of consumer behaviour, the degree of
external information search is affected by the perceived risks associated with both the product/or
service to be purchased (e.g. its cost), and the costs of the information search itself in terms of
time, money and psychological cost. Heuristics and satisficing behaviour may be employed to
reduce the costs of external information search. The third stage of the decision-making process
- alternative evaluation – is typically assumed to involve the selection of one choice from
alternatives. However, information may also be sought even when alternatives are not available
or not considered; in this case, information may serve a confirmatory purpose, reduce anxiety or
enhance the travel experience (van der Horst, 2006). This point was informative for the
development of the current research by helping to conceptualise informal information not just as
a factor in the making of active travel choices, but also as something which might play a wider
social and psychological role. The apparent complexity of this role indicated that a qualitative
research methodology might provide an appropriate means of exploring this area.
One key assumption of information processing theory and other “grand models” of consumer
behaviour is that behaviour is individual (Gilbert, 1991), an assertion criticized by Sirakaya and
Woodside (2005) in their review of decision-making models in the tourism literature. These
authors comment that all the models they reviewed “accept that other individuals affect the
decision-maker, but do not address active interaction with other individuals or sources along the
decision-making process” (p.829); most research focuses on individual psychological, not social
variables. They therefore argue that a different theoretical approach is needed to accommodate
Literature Review
the fact that recreational travel is often a social activity involving family and friends, who may be
significant mediators of trip decisions, particularly destination. Although the everyday trip context
on which the current study focuses differs from tourist travel in the sense that it may be less
frequently undertaken in a family or friendship group, Sirakaya and Woodside‟s (2005)
conclusions concur with observations from the transport literature that the social context of travel
information-use has been neglected (e.g. Lyons et al., 2007). For example, in their 1997 review
paper, Jackson and Polak identify a number of elements which contribute to behavioural
responses to travel information, distinguishing between factors located within the individual, and
factors located in the environment. Among the latter they cite “expected norms of behaviour:
societal constraints and expected patterns of behaviour associated with a particular
environment” (p21). However, this dimension is not considered in detail, nor does it feature
explicitly in their corresponding model of the decision-making process.
Choice set models
One of the models in the tourism literature which does emphasise the role of other people in the
travel decision process, albeit in the sense that others may influence an individual‟s decision,
rather than placing social interactions at the centre of the decision process, is Um and
Crompton‟s model of the “pleasure destination choice process” (1990). Adopting a choice set
approach, these authors conceptualise (tourist) travel decision-making as a funnelling process,
in which alternative travel options are eliminated until a final decision is reached. The three
choice sets identified in this model are the awareness set (all the alternatives of which the
traveller is aware), the evoked set (a smaller set of probable choices) and the final decision
selection. This model was developed to explain the process of leisure destination choice; it could
be argued that in these circumstances a wider range of choices is available than would typically
be the case for the more day-to-day trip decisions on which the present study focuses. However,
the model includes some interesting relationships between concepts which are also applicable to
the role of word-of-mouth in the everyday travel context. It integrates the role of social
interactions into the decision process at both a “passive information catching” level and an active
decision-making level, as well as incorporating social-psychological constructs, informed, for
example, by Ajzen‟s (1991) work on attitudes and behaviour (see Section 2.2).
Figure 2-1 shows an adapted version of Um and Crompton‟s model (1990). In the model,
external inputs contribute, through a process of belief formation, to the creation of an awareness
set, prior to the initiation of any decision process. Although in this case the awareness set
comprises all possible “pleasure travel destinations”, it might also apply to all possible travel
alternatives for different types of trip, be they matters of mode, route, travel time or destination.
Literature Review
The external inputs comprise personal past experience of the trip (more typically categorised as
an internal factor in other models), words and images disseminated through the media, and
awareness of trip attributes absorbed through face-to-face interactions with other people,
including other people communicating their direct or indirect travel experiences. These external
stimuli contribute first to the awareness set through a process of “passive information catching”,
and later to the evaluation of the evoked set – the narrowed down set of alternatives, through a
process of active information search, once the decision process has been initiated as a
consequence of the need or desire to make a trip. Beliefs about trip attributes are formed by
exposure to external stimuli, but are also shaped by individual factors such as sociodemographics, lifestyle, personality, motives, values and attitudes. Internal and external inputs
are integrated through a cognitive process into the awareness set and the evoked set of
alternatives. Situational factors such as time and money are considered in the process of
narrowing alternatives to form the evoked set, from which the final choice will be made. This
model is of particular interest because it includes the role of interactions with others not just as a
form of active information search, but also as a factor in the formation of beliefs though passive
information absorption - this occurs before the initiation of the choice process.
Figure 2-1: Model of the travel choice destination process, adapted from Um and
Crompton, 1990.
External inputs
Cognitive constructs
Belief f ormation
Internal inputs
(Passive inf ormation catching)
Past personal
Initiation of Choice
Evolution of an Evoked Set
Face-to-f ace
interactions with
(Consideration of
situational constraints)
Belief f ormation
(Active inf ormation searching)
Evoked Set
Destination selection
Literature Review
The desire for informal types of travel information
Research into the types of travel information which people need (or would like to have) has
demonstrated the importance of accurate „formal‟ information such as journey time, costs, and
route planning (Social Research Associates, 2004); these have tended to constitute the core
characteristics of traveller information services. Referring back to Figure 1-2, p.8, they are
located within the upper quadrants (formal information content). However, studies on user needs
have found that respondents consider information on reliability to be even more important
(Transportation Research Group, 2000; MORI, 2001, Accent, 2002). While it is possible to
measure and present data on, for example, the average adherence to schedule of a particular
service, notions of „reliability‟ are subjective, and here the boundary between formal and informal
information becomes less clear. Would a potential bus user be more influenced by data
presented by the bus company, or by a friend‟s perception that the service is „unreliable‟? The
TRG study (2000) found information on „convenience‟ to be the third most important facet of a
travel information service (more important than information on cost), which is an area of even
greater subjectivity. Chorus et al. (2006a, 2006b) cite a number of other empirical studies which
suggest that travellers may want advanced traveller information systems to provide information
about „soft‟ characteristics of travel alternatives, such as convenience and „image‟. These
factors are located in the lower quadrants of Figure 1-2 (informal content).
SRA (2004) explored convenience factors relating to interchanges and end-legs, suggesting that
difficulties with physical accessibility and direction-finding at interchanges may be a barrier to
travelling by public transport for some people (especially those with a walking disability, or
travelling with children or heavy luggage). This type of information is difficult to provide in a topdown way, because “what one user wants or needs in the way of “convenience”, another does
not…. “ (SRA, 2004, p23), but it may be a “killer issue” for some people in deciding how to travel.
This might particularly be the case for those who fall outside the socio-demographic group most
commonly using advanced traveller information - the male, able-bodied, well-educated business
traveller (Chorus, 2006b). For example, research for Transport Direct concerning dyslexic
travellers found that information on end-legs and interchanges was important, if not essential, for
them to embark on public transport journeys (Lamont and Lyons, 2007). Other concerns which
are crucial for some people are fears about personal security, likelihood of getting a seat, and
availability of refreshments and toilets (SRA, 2004). While some of this information is „factual‟,
concerns such as personal security have a strong emotional and psychological dimension, which
are unlikely to be satisfactorily addressed through formal information provision.
Literature Review
The use of word-of-mouth travel information sources
What research evidence exists to suggest that people acquire travel information through wordof-mouth, as well as, or instead of formal sources (the right hand quadrants in Figure 1-2)?
Results from the 2005 and 2006 UK National Travel Survey show that 24% of those who had
sought travel information for a public transport journey during the previous six months had
“asked a friend”, whilst 28% of those who had sought information to plan a car journey had done
the same. This compares with the 24% who had consulted Transport Direct or another website
to plan a public transport journey, and 31% who had used one of these sources to plan a car
journey. It is not possible to tell from the data how different information sources had been used
in combination. Qualitative research by Farag and Lyons (2008a) showed that the use of wordof-mouth travel information sources, such as family, friends and colleagues, varied depending on
how trustworthy the information obtained was perceived to be. Not surprisingly, participants were
more likely to consult informal sources if they knew people who lived in, or frequently visited,
their trip destination. With regard to specific groups of people, research has shown that those
with learning difficulties tend to rely on word-of-mouth advice when planning journeys (TTR,
2004). Research in Australia found that users of ATIS also sought advice from family, friends
and colleagues (Karl and Bechervaise, 2003). However, the role of word-of-mouth in influencing
journey planning decisions has not been widely explored in the transport literature.
The role of word-of-mouth information has, however, been studied more extensively in the
tourism and consumer studies literature (e.g. Gretzel et al., 2007, Murphy et al., 2007, Bieger
and Laesser et al., 2004). Sirakaya and Woodside (2005) suggest that because the perceived
financial and emotional risks associated with holiday decisions are often high, “word-of-mouth or
personal information sources are more influential than impersonal media sources in decisions”
(p.826 ). Referring to early research on the same subject, Gitelson and Kerstetter (1994) found
that friends and relatives were usually cited as the most frequent and most credible source of
pre-trip information. Um and Crompton (1990) cite a number of earlier studies from the tourism
literature highlighting the importance of travel information received from family, friends and other
social contacts. For example, Gitelson and Crompton (1983) found that 74% of respondents had
received travel information from friends and family, whilst Walter and Tong (1977) reported that
family and friends were the most influential source of information for destination choice.
The above studies concerned face-to-face word-of-mouth. However the phenomenon of
“electronic word-of-mouth” via communications technologies is receiving growing research
attention across a number of disciplines. For example, in management science, Dellarocas
(2003) has explored the building of trust and fostering of cooperation through online feedback
mechanisms, using a game theory approach. In health research, Ziebland et al. (2004)
Literature Review
undertook a qualitative study of the use of the internet by cancer patients; one of the main uses
of the internet for these patients was found to be the seeking of experiential information from
other patients. Closer to the field of this PhD, Chabot (2007) found „Web 2.0‟ tools to be exerting
a growing influence on tourists‟ decision-making. In a survey of users of Trip Adviser holiday
reviews (www.tripadviser.co.uk), Gretzel et al. (2007) found that users were affected by social
cues about the writers of reviews when evaluating content. Motivations for posting reviews
included concern for other consumers, and “need for positive self-enhancement“ (relevant, within
this thesis, to the question of why people may provide travel information to others). Literature on
online word-of-mouth in areas other than transport will be considered further in Section 5.5.
This section has highlighted existing theoretical approaches to (travel) information-use, which
largely treat this as an individual rather than a social process. It has also summarized some of
the references in empirical studies to informal types of information and word-of-mouth sources
(including electronic communication), the majority in the tourism rather than the transport
literature. It is postulated that in the absence of serious consideration of social factors within
conventional decision theory, it would be fruitful to turn to social-psychological theories of
behaviour in order to seek a greater understanding of the factors which might influence informal
information-use and its diffusion through word-of-mouth.
Social-psychological theories relevant to word-of-mouth information processes
Together with theories of decision-making, social-psychological theories of behaviour
contributed to the development of a conceptual framework for this project. These theories
helped to identify some of the key concepts and mechanisms underlying the social transfer of
travel information, such as social norms, pro-social values, self-concept, social identity and trust.
A review was undertaken to find out which, if any, theories have been applied to social
dimensions of information behaviour (and which, if any, had been applied to travel information in
particular), and to identify any gaps in knowledge. This part of the review was used, in particular,
to help formulate and answer Research Question 3 (how social-psychological factors might
influence the use and effects of word-of-mouth information). Later, during the design of the
applied research (Phase 2), some of the theories reviewed in this and the following section (2.3)
were used to refine sub-question 4.4 on the role of social-psychological factors in the context of
information-sharing via a web-based traveller information system.
As an over-arching theory, Bandura‟s social learning theory (1977) provides a helpful starting
point for the understanding of word-of-mouth information processes, by explaining human
behaviour in terms of a continuous reciprocal interaction between personal/behavioural and
environmental determinants. It contends that people learn not only from their own direct
Literature Review
experiences, but also from the experiences of people around them. Vicarious experience of
others‟ behaviour has as much influence on peoples‟ behavioural choices as do their own direct
experiences. Sharing experiences through word-of-mouth is one of the direct ways in which
social learning occurs, although the learning process may also occur non-verbally through the
observation of others. In relation to travel behaviour, Sunitiyoso (2008) investigated the impact of
social learning in small groups and its implications for travel behaviour modelling. Laboratory
experiments within the context of a transport social dilemma (see below) revealed that people
may be influenced by the behaviour of others when they have access to social information about
other people‟s decisions.
A number of specific constructs relevant to the transfer of social information appear within
expectancy-value attitude theories, such as the Theory of Interpersonal Behaviour (TIB)
(Triandis, 1977), and the Theory of Planned Behaviour (TPB) (Ajzen,1991). The TPB proposes
that behavioural intentions are partly influenced by subjective norms: the perception that
“significant others” would or would not approve of the behaviour in question. This construct was
measured using a travel behaviour survey conducted among UWE students by Sunitiyoso
(2008). The survey showed that 49% of respondents considered it important that family
members approve of their mode of travel to the university. This type of normative information
about „approved‟ travel behaviours is likely to be transmitted through word-of-mouth amongst
close friends and family. The TIB adds complexity to the TPB by separating the „social factors‟
which influence intentions and behaviours into norms, roles and self-concept (see Figure 2-2).
Social norms are rules about what does and does not constitute acceptable behaviour; roles are
“sets of behaviours that are considered appropriate for persons holding particular positions in a
group” (Triandis, 1977, quoted in Jackson, 2005), whilst self-concept refers to a person‟s
perception of him or herself (Anable, 2006). Self-concept is linked to Social Identity Theory,
which suggests that people‟s sense of identity derives from their membership of certain social
groups which have an emotional and value significance for them (Tajfel, 1982). Aspects of the
work of Tajfel and colleagues in relation to social influence within the group will be reviewed in
Section 2.3. Social norms, roles and self-concept are all factors which might affect the impact of
word-of-mouth information on individual recipients; for example, information might be perceived
as more reliable if it emanates from a person with whom the recipient identifies, or if it reflects a
norm of behaviour within a reference group.
Despite their inclusion of social factors, expectancy-value theories such as the TIB and TPB are
better known for their treatment of the attitude construct. An attitude is defined as a
psychological tendency which is expressed by evaluating a particular entity with some degree of
favour or disfavour (Eagly and Kulesa, 1997). These theories posit that an attitude is an
Literature Review
evaluative response to an attitude object arising from beliefs about the characteristics of that
object, weighted by the value an individual attributes to it. Beliefs can be said to represent the
information people have about the world around them, which they internalise as knowledge, be
this objective (factual) knowledge, or a subjective assessment of their own knowledge. Hence,
knowledge might be considered a precondition to attitude. Theories such as the TPB and TIB
treat the formation of attitude as a cognitive evaluative response to an attitude object, but
attitudes can also arise from an affective evaluation of an attitude object based on the positive or
negative emotions which it induces (Lyons et al., 2008, Bohner, 2001). Whilst the TIB identifies
affect as an antecedent to behavioural intention, it treats this as a separate factor rather than
one which contributes first to attitude formation (see Figure 2-2). An understanding of constructs
such as belief, attitude, intention and affect was important for this research because they may all
be subject to influence through word-of-mouth. For example, attitudes towards a mode of
transport might be influenced by word-of-mouth information, whilst social factors such as norms,
roles and self-concept may help to identify reasons why such influence occurs.
Figure 2-2: Triandis‟ Theory of Interpersonal Behaviour (source: Jackson, 2005)
More detailed insights into the concept of social norms can be found in the work of Cialdini et al.
(1990), who differentiate between descriptive social norms - “what most people do”; and
injunctive social norms –“ what ought to be done”, according to the moral rules of the social
Literature Review
group. Norms also constitute an important part of Schwartz‟s Norm Activation Theory (Schwartz,
1977), although in this case they are personal norms, as opposed to social norms. Personal
norms are defined as self-expectations for specific action in particular situations, arising from a
feeling of responsibility for others, and are used by Schwartz to explain pro-social behaviour
(Schwartz, 1977). There is extensive use within the transport literature of Norm Activation
Theory (NAT) and other models incorporating concepts of social responsibility and moral norms,
particularly where travel behaviour is placed within a framework of pro-environmental behaviour
(e.g. Taniguchi et al., 2003). Anable et al. (2006) remark that evidence of the role of social
norms in travel behaviour is mixed, citing a number of contradictory findings from empirical
studies comparing the effects of social norms alongside factors such as personal norms, moral
norms and habit (e.g. Bamberg and Schmidt, 2003). Anable et al. (2006) conclude that “there is
evidence of various social influences on travel and it would appear that social norms merit
distinct and thorough examination alongside personal norms” (p.108).
Many components of earlier social-psychological theories have recently been incorporated into
the Comprehensive Model of Consumer Action (Bagozzi et al., 2002). Subjective norms and
social identity are included within the many other influences on consumer behaviour within this
composite model. In the earlier Theory of Trying, Bagozzi had argued that many consumer
behaviours can be studied from the perspective of “trying to act”, hence these behaviours are
goal-directed (Perugini and Connor, 2000). Bagozzi‟s model may be relevant to this study
because the acquisition of travel information can be understood as a goal-directed behaviour:
information may help to achieve the practical goal of making a journey (Farag and Lyons,
A further group of social-psychological theories, such as Kelley and Thibaut‟s Interdependence
Theory (1978) explain some social behaviour in terms of cooperation and competition amongst
interacting individuals. They are of interest to this thesis because they concern social interaction,
through which word-of-mouth information is transmitted. These theories have been applied in the
field of travel behaviour with regard to modal choice (e.g. van Vugt et al., 1995), but without
specific reference to information behaviour. Interdependence Theory proposes that
interdependent persons may find it mutually beneficial to perform a pro-social transformation, in
which each person starts to take decisions on the basis of what benefits the other person, rather
than him or herself. One of the factors determining whether or not such a transformation takes
place is social value orientation (McClintock, 1972). McClintock identifies four main categories of
disposition: altruists and cooperators (both pro-social); and individualists and competitors (both
pro-self). This approach is relevant to the proposed research because social values might affect
people‟s willingness (or lack of it) to share travel information with others. Moreover, pro-social or
Literature Review
pro-self values might be reinforced, and corresponding behaviours encouraged through
interaction with others. In social psychology, values are held to be beliefs relating to desirable
end states which transcend specific situations, guide selection or evaluation of behaviour,
people and events, and are ordered by relative importance (Schwartz, 1994, cited in Hewstone
and Stroebe, 2001). Pro-social values help to explain pro-social behaviour, which may be a
factor in people‟s motivation for providing information to others.
There are a number of theories which may be used in combination to understand why people
help one another: biological; individualistic; social-systems; and interpersonal theories (Bierhoff,
2001). The biological, or evolutionary psychology approach, suggests that pro-social behaviour
is genetically determined, and can be explained by the evolutionary advantages of both kin
selection, and reciprocal altruism among non-relatives. The individualistic approach also focuses
on individual tendencies, but argues that pro-social behaviour is not necessarily determined by
our genes, but may be acquired through social learning (Bandura, 1977). According to the
individualistic approach, people may be more helpful at some times than at others, depending on
their mood, or they may be more consistently helpful because of pro-social personality traits,
which include empathy and social responsibility. Therefore, some people might consistently wish
to offer helpful information, whilst for others, this might be determined by situation or mood. The
social systems approach is also concerned with social learning, but emphasises the influence of
factors inherent within a social system, such as cultural norms and rituals shared within a
community (e.g. norms of information-sharing). Finally, the interpersonal approach stresses the
importance of interdependence structures, in which outcomes for the individual are influenced by
the actions of others. It is associated with interdependence theory (Thibaut and Kelley,1978),
which was outlined in the previous paragraph.
Interdependence theory has provided a conceptual framework for psychology research on prosocial behaviour using experimental and stated preference methods. In these studies,
participants were generally presented with a social dilemma – a situation where the immediate
interests of the individual conflict with the long-term collective interest - to see whether they
would exhibit cooperative or competitive behaviour. Some of these studies were framed within a
transport context, often relating to modal choice, as the choice between car and public
transport/cycling/walking may be regarded as an archetypal social dilemma (van Vugt et al.,
1995, van Vugt, 1997, van Lange et al. 1998). The review of literature in this field did not reveal
any application of the theories to the social transfer of travel information, but produced a number
of relevant insights into the relationship between pro-social values and cooperative behaviour
(e.g. Liebrand, 1984); social interaction and cooperative behaviour (e.g. Sunitiyoso, 2008,
Avineri, 2006); and social values and trust (van Lange et al. 1998). The research identified in this
Literature Review
area was exclusively quantitative, much of it using laboratory experiments, which do not have
high ecological validity. In-depth, qualitative study of these concepts in a real-world context
might therefore improve understanding in this field.
The theories of social interaction outlined in the preceding two paragraphs were the only
theories in this review which incorporated the concept of interpersonal trust. If word-of-mouth
information is to have an influence on belief, attitudes and behaviour, trust in both the
information-giver and the information itself are likely to be important factors. In interdependence
theory, trust is articulated as trust in others to cooperate in a situation of interdependence, such
as a social dilemma. However, this is not directly applicable to the case of seeking or giving
information, as the participants in an interaction are not necessarily in a situation where the
outcome for each depends on the behaviour of the other (unless the giving of information is
interpreted as being motivated by long term reciprocal altruism). Although a „theory of trust‟ was
not identified in the carrying out of this review, the concept of interpersonal trust has been
researched in a range of disciplines, notably economics, psychology and sociology. In a crossdisciplinary review, Rousseau et al. (1998) argue for a broad definition of trust as : “a
psychological state comprising the intention to accept vulnerability based upon positive
expectations of the intentions or behaviour of another” (p.395). Trust arises in conditions where
both uncertainty and (inter-) dependence are present, although, as suggested above,
dependence is unlikely to be mutual in the case where one person provides information and the
other receives it. Trust may take different forms, but two conceptualisations which are pertinent
to the present study are calculus-based trust and relational trust. The former is based on rational
choice and is typical of short-term (often economic) interactions where the truster must calculate
whether the trustee intends to behave in a way which is beneficial to the truster. Trust is derived
in part from credible information regarding the intentions or competence of the other person. In
contrast, relational trust derives from repeated interactions over time between the two parties,
where trust is based on information available to the truster from within the relationship itself,
within which emotion may play a part (Rousseau et al.,1998). Within social psychology,
distinctions have been drawn between reliability and emotional trust. For example, JohnsonGeorge and Swap (1982) found that their experimental subjects could readily attribute reliability
to another subject whom they had met during the experiment, but emotional trustworthiness was
more difficult to determine in laboratory conditions. In social psychology, research on relational
trust has often focussed on trust in a specific relational - often romantic - partner, but was
broadened by Couch and Jones (1997) to include trust in one‟s social network of family and
friends, termed by these authors as network trust. This form of relational trust was of particular
interest to the current research in terms of the trust placed in information provided by a person
within one‟s social network.
Literature Review
Whereas the use of word-of-mouth information is likely to be affected by the degree of
interpersonal trust inherent in a person-to-person interaction, trust in formal travel information
could be described as a form of institutional trust (Rousseau et al., 1998). Anable et al. (2006)
consider the role of trust in relation to the official institutions seeking to motivate support for
climate change mitigation. Although their review failed to reveal any specific evidence in relation
to trust in information and government action on climate change and transport, they cite a
number of studies in other fields which demonstrate relative differences in trust inspired by
different information sources. National government, the EU and relevant businesses tended to
inspire less trust than non-governmental organisations or scientists independent of government
or industry. It was of interest to explore whether this distrust in „official‟ information sources might
be mirrored in the field of travel information, and how far word-of-mouth information might be
considered trustworthy in comparison with these sources.
This part of the review has revealed a number of social-psychological constructs and theories,
such as social learning, social norms, social identity, pro-social values and trust, which might be
explored in the context of informal information and word-of-mouth processes. These constructs
and theories may help to explain some of the reasons why people seek information from, and
offer it to, other people, and the extent to which it might influence them. Very little literature was
identified where such theories had been applied to (social) travel information-use, suggesting a
novel theoretical approach for this thesis. Figure 2-3 provides a summary of the main constructs
and theories which arose from the literature review and which were to be investigated within the
PhD research as factors which might contribute to processes of social influence when
information is obtained or offered through word-of-mouth. First, however, more detailed
consideration is given to theories of social influence.
Social influence and the social psychology of the group
Of particular interest for the second empirical phase of the research were theories of social
influence, dating back to the experimental social psychology of the 1950s, and theoretical work
on the social psychology of the group, developed by Tajfel, Turner and colleagues in the 1970s
and 1980s. Both the group context of the Phase 2 research (use of an experimental case-study
information system by a small group), and the interest in word-of-mouth as a channel for social
influence, made two related areas of theory particularly apposite: the dual process theory of
social influence (Deutsch and Gerard, 1955), and self-categorisation theory (Turner et al., 1987).
Deutsch and Gerard (1955) reinterpreted some of the „classic‟ experimental studies of social
influence of the 1930s to 1950s by differentiating between informational and normative social
influence. Self-categorisation theory, a development of Tajfel and Turner‟s (1986) social identity
Literature Review
theory, adds the concept of referent informational influence to those of normative and
informational influence. It may be useful to clarify at this point that, in drawing on this area of
social-psychological theory, the term information is used in its broadest sense – that is, raw data
which require interpretation in order to derive meaning (Floridi, 2010). Thus, the term is used in
this thesis to encompass the diverse forms of information, both „factual‟ (for example, “the cycle
path starts here”) and „social‟ (for example, “people like us cycle to work”) which are
communicated through social interaction, and not just the factual “semantic information” such as
one finds in a railway timetable (Floridi, 2010). The latter definition is the one which is usually
associated with “travel information”, so it is perhaps unsurprising that the theories discussed in
this section have rarely been used in the study of travel information. Exceptions can be found in
the fields of tourism, business and information systems research; for example, the concept of
informational and normative influence informed recent studies of online (leisure) travel
information-use by Arsal et al. (2010); Casaló et al. (2011), and Mendes-Fihlo and Tan (2009).
Dholakia et al. (2004) built a social identity variable into their social influence model of consumer
participation in virtual communities. Cheung et al. (2009) categorised the consistency and rating
of online recommendations as normative determinants of information credibility, and found these
factors to be influential alongside informational determinants such as argument strength and
confirmation of prior belief.
According to Deutsch and Gerard‟s dual process theory (1955), informational influence is based
on the acceptance of information obtained from others as evidence about reality, whereas
normative influence is based on the need to conform with the positive expectations of others,
particularly in a group environment. Both processes may operate in parallel, although the relative
importance of each will vary according to the situation. The former process reflects a
dependence on others for the reduction of uncertainty, whilst the latter reflects the need for
social rewards such as acceptance and approval. In the field of word-of-mouth travel
information, an individual might accept information from cyclists about the lighting levels on a
particular cycle route as evidence of reality (informational influence), because these cyclists
have experience of using the route after dark, so their opinion is to be trusted. However, they
may also be subject to a more subtle normative influence – that it is quite „normal‟ behaviour
within this group of cyclists to use this route after dark.
Informational influence is associated with a private acceptance of, and trust in, others‟ opinions
(conversion), whereas normative influence is believed to encourage public conformity
(compliance) without an internalised change to an individual‟s private attitudes. A strong
normative influence, as conceived by Deutsch and Gerard (1955) might therefore, in certain
situations, help to explain why expressed intentions to change behaviour, such as switching from
Literature Review
driving to cycling to work, may not materialise into actual behaviour change (the “attitudebehaviour gap”; e.g. Ajzen and Fischbein, 1977), or may affect only a temporary change. Hence
social psychologists have tended to regard only informational influence (within this specific
conceptualisation) as „true influence‟ (Turner et al, 1987). This is consistent with concepts of
impression management and self-presentation, which suggest that people comply with group
norms in order to create a positive self-image in social interactions (e.g. Leary, 1995).
The experimental work of Deutsch and Gerard and other early social influence researchers
elucidated specific social processes in a group context, but Tajfel, Turner and colleagues were
interested in what happens to people‟s identity in group settings; they argued that in such
settings people‟s psychological processes are qualitatively transformed (Wetherell, 1996), as
personal identity gives way to social identity through a process of “depersonalisation”. Whilst still
maintaining their identity as unique individuals in interpersonal comparisons, people can also
perceive themselves as members of a social group with the characteristics of that group, and
may modify their attitudes and behaviour to comply with norms within the ingroup (reference
group). Perceptions of group membership are fluid, allowing an individual to categorise him or
herself as a member of, and identify with, different groups, at different levels of abstraction, as
they become more or less salient. Self-categorisation theory is thus a general theory of group
behaviour which emphasises the effect of self-definitions (self-stereotypes) in the context of
social groups. Thus, an individual may categorise him or herself, for example, as a man or
woman, a student, a parent, a car-driver or a cyclist at different times in different circumstances,
and may alter his or her behaviour depending on the saliency of a particular social identity.
Turner identified a form of social influence called referent informational influence, whereby
people adjust their identity, attitudes and behaviour to correspond with the collectively defined
attributes of their social groups (Wetherell, 1996). He argued that normative and informational
influence were not as easily distinguishable as Deutsch and Gerard‟s theory (1955) suggested,
and that referent informational influence integrated both concepts: the basic influence process is
one where the normative position of people categorised as similar to self tends to be subjectively
accepted as valid (Turner, 1991). Thus, it is not the informational content per se of others‟
opinions and actions which matters, but the extent to which it is validated by ingroup consensus
(Turner et al. 1987). So, returning to the earlier example of shared information about lighting on
a cycle path, Turner‟s theory would suggest that consensus amongst a “reference group” of
cyclists (e.g. work colleagues) about the safety of using the path after dark would exert more
influence on an information-seeker within the same group than factual content about the lighting
itself. Both dimensions of the information would be deemed more trustworthy than information
provided by an outgroup (for example an unidentified cyclist or information from an „official‟
Literature Review
source such as the local council). However, this interpretation might be questioned as one which
underestimates individual differences in personality and behaviour, particularly in terms of
susceptibility to group influence, and is especially incongruent with the concept of the „individual
rational decision-maker‟, within which the design and study of traveller information systems has
traditionally been framed. The role of word-of-mouth as a channel of referent social influence
alongside more conventional notions of individual, rational information-use, helped to provide a
novel theoretical approach within this thesis.
Figure 2-3 summarises the key social constructs and theories from this part of the literature
which were thought at this stage to offer some explanation as to how word-of-mouth travel
information might influence beliefs, attitudes, intentions and behaviour (if evidence of such an
influence were to be found). In the subsequent research design, the individual constructs - to the
left of the figure – were to inform the design of interview questions about social-psychological
processes, as well as guide the interpretation of the findings. The over-arching theories, to the
right of the figure, were thought to offer insights into the overall influence process which might
assist in the interpretation of findings.
Literature Review
Figure 2-3: Contribution of social psychology theories and constructs to understanding
of the topic
Acquisition of travel information
Obtained/offered through
Possible contributory
Possible contributory
Individual constructs
from theory
Social psychology theories
Social learning
(Bandura, 1977)
Social norms
Personal norms
Dual process theory of social
(Deutsch and Gerard, 1955)
Social identity
Self-categorisation theory
(Turner et al., 1987)
Interdependence theory
(Thibaut and Kelley, 1978)
Pro-social values/behaviour
Influence on travel decision process?
(beliefs, attitudes, intentions, behaviours)
Literature Review
ICT, social networks and travel
Another relevant body of knowledge which might inform our understanding of ways in which
travel information is shared, and how word-of-mouth information might influence travel decisions,
is sociological theory and empirical research addressing social networks and information
diffusion. Two areas of particular interest are social network theory and the diffusion of
innovations (Rogers, 2003).
Social network theory views social relationships in terms of nodes and ties; nodes are the
individual actors within the networks, and ties are the relationships between the actors. Mark
Granovetter (1973) developed the influential “Strength of Weak Ties” theory, which suggests that
the weak ties connecting acquaintances provide a bridge between “clumps” of closer personal
networks (strong ties). Information and innovation spread more effectively through weak ties
than strong ones, as networks of close friends and family are likely to have the same information
already. Social network analysis has recently gained some importance in transport research, not
least in the field of activity-based transport modelling, where much work is being undertaken to
combine spatial and social network models (cf. Frontiers in Transportation conference,
Amsterdam, October 2007). It is suggested that by studying social networks we can gain insights
into the generation of social activities, which might, in turn, contribute to improving the
behavioural aspects of agent-based activity demand models (e.g. Carrasco and Miller, 2006).
However, the emphasis of such research on the quantitative and geo-spatial aspects of social
interactions means that little attention has been paid to understanding the quality of different
relationships within a person‟s social network, and the way that the different social dynamics
within interactions with different people may lead to differing patterns of information diffusion,
social influence and travel choice behaviour. It is suggested by Kozinets (2010) that there are
possible synergies between the structural analysis of social networks and more meaning-centred
qualitative approaches, in particular through online ethnography. Here he refers to the method of
“netnography", which will be discussed in Section 5.5. Netnography can, for example, help to
provide explanations for the structural characteristics such as power and influence in
relationships and social ties, which are uncovered by social network analysis.
Little is known about the reasons why information from particular people is given more weight
and inspires greater trust (and therefore why some sources have a greater impact on other
people‟s attitudes and travel behaviour). Kozinets (2010) identifies this as an area of knowledge
to which social network analysis might contribute (in combination with qualitative methods),
suggesting a practical application which is highly pertinent to the current study: “Trusting
relationships, linked to strong ties, are also relevant to understanding and planning the online
provision of many types of public information” (p.53).
Literature Review
Kozuki and Mahmassani (2009) have developed an interesting model of the dynamics of
information acquisition, social interaction and opinion formation in the context of activity and
travel choice behaviour, in which they simulate the effects of word-of-mouth mechanisms on the
formation and propagation of opinions about a new transport service. However, this model
makes a range of assumptions about human behaviour which may benefit from greater
exploration. This suggests that qualitative research may complement the more typical
quantitative approaches by delving beneath some of the assumptions made about human
behaviour within modelling approaches.
The positive impact of socialising with one‟s neighbours and local friends is demonstrated by
empirical results from a 20 year longitudinal study in the United States, which found that people
were more likely to describe themselves as happy if they were connected to other happy people
– this emotional state seemed to spread through connections within social networks up to “three
degrees of separation” (Fowler and Christakis, 2009). However, this effect required the presence
of physical proximity: friends living within a half mile radius had the greatest positive effect, but
this effect declined with distance. The second most positive effect arose from the presence of
happy next-door neighbours, whilst other neighbours living in the same block had no effect.
A social network approach is of relevance to the study of word-of-mouth information diffusion in
terms of the networks through which people share travel information. In addition to the „higher
level‟ sociological application of social network theory, described in the preceding paragraphs, it
has been applied to travel information aquired through social interaction in the context of tourism
and leisure travel. For example, Axup and Viller (2006) used a social network analysis to study
social interactions among backpackers in Australia, in order to explore the potential for mobile
social software (MoSoSo) specifically designed to facilitate the exchange of (tourist) travel
information. Their studies involved pairing up backpackers with similar past or present travel
experiences or future travel plans, to see which type of interactions were valued most highly.
Among their results was the finding that participants rated the information they received from the
other person more highly if they shared similar personality traits and values. The Axup and Viller
research is especially relevant to this project because it explores processes of social interaction
among travellers, in order to inform the development of communications technologies to facilitate
information exchange. The major difference is that their research concerns leisure travel and is
largely concerned with information-sharing about destinations rather than journeys per se.
Finally, Diffusion of Innovations theory (Rogers, 2003) defines diffusion as a process in which an
innovation is communicated through certain channels over time among members of a social
system. This theory could be applied to the use of traveller information because information is
part of the “innovation decision process”, where the innovation in question is changed travel
Literature Review
behaviour. For example, cycling to work might be considered innovative in an environment
where everyone else commutes by car, even if there is nothing innovative about riding a bike per
se. “The innovation-decision process is essentially an information-seeking and information–
processing activity in which an individual is motivated to reduce uncertainty about the
advantages and disadvantages of the innovation” (Rogers, 2003, p.14). In this sense, seeking
traveller information is one step in the process of deciding whether or not to adopt a new travel
behaviour (Halpern et al., 2004). This approach is therefore relevant to the study, within this
project, of word-of-mouth information diffusion. No previous research was identified which used
qualitative methods to explore the diffusion of travel behaviours through the medium of word-ofmouth information.
The study of “online communities”
This final part of the literature review outlines some of the recent research into online culture
(“netnography”), as well as some of the studies in a range of fields – such as information
science, business studies and education – which have sought to improve understanding of
social factors in the functioning of online communities, sometimes through the application of
social-psychological theory. Examples from this area of literature informed both the design of the
Phase 2 empirical work, and the interpretation of the results. Those studies which make
recommendations for the design of „social‟ features in order to improve the way in which online
communities function are of particular relevance to the final research question in this thesis,
which concerns the possible incorporation of „social design features‟ into advanced traveller
information systems.
A set of methodological guidelines is provided by Kozinets (2010) for the study of the social
interaction which occurs through the internet and related information and communications
technologies. Kozinets argues for the introduction of the neologism netnography, which is
different from conventional ethnography, since online social experiences differ from face-to-face
social experiences. Hence, the experience of studying virtual communities is different from that
of studying physical communities. The term "virtual community" was originally coined by the
internet pioneer Howard Rheingold (1993). Rheingold defined virtual communities as: "social
aggregations that emerge from the net when enough people carry on…. public discussions long
enough, with sufficient human feeling, to form webs of personal relationships in cyberspace".
(Rheingold, 1993, p.5). It was of interest for Phase 2 to find out whether an online space for
informal information-sharing about local travel might evolve into a „community‟ of this sort.
Kozinets (2010) argues that an online (or virtual) community comprises "group of people who
share social interaction, social ties, and a common interactional format, location, or "space"
Literature Review
(p10). He also suggests that the term „community‟ requires sustained social interaction, and a
sense of familiarity, or shared identity, between members of the group.
“Netnography” is just one of the methods used in the literature for the study of existing online
communities (i.e. „realworld‟ communities which were not set up purely for research purposes).
For example, Zhang and Watts (2008) investigated a Chinese online (leisure) travel community
from a perspective of communities of practice, and considered applications to business
development: how organisations can better utilise online social structures for their knowledge
management practice. Cheung et al. (2009) conducted a survey of users of a popular consumer
forum in China. As outlined in Section 2.3, these authors applied the dual process theory of
social influence (Deutsch and Gerard, 1955) to measure normative and informational-based
determinants of the perceived credibility of word-of-mouth consumer recommendations. Hall and
Graham (2004) studied the ways in which members of an online community - a code-breaking
Yahoo group entitled CipherChallenge - collaborated in the pursuit of a single financial prize. It
showed that the initial motivation for people to join the group was to gain information for their
personal benefit, but over time, the online interactions invoked a desire for members to
reciprocate the help they had received. This was of relevance to the study of reciprocal helping
(pro-social behaviour) within the Phase 2 case-study for this thesis.
Other studies have involved the setting up of an online environment for research purposes (e.g.
Tu and McIsaac, 2002; Wilson et al., 2006: Hall and Widén-Wulff, 2008) or the setting up of
experimental groups within an existing online community (e.g. Ling et al., 2005). New online
social spaces were found to have been set up most frequently in the context of online learning.
Tu and McIsaac (2002) used mixed methods to study "social presence" in the context of learning
- this is defined as a measure of the feeling of community that a learner experiences in an online
environment. One of the findings of the qualitative part of their study was that more variables
contribute to “social presence” than is suggested by the relevant theory. This observation
supports the view held in this thesis that qualitative research can be useful for building up a
more detailed picture of some of the constructs within social-psychological theory. Hall and
Widén-Wulff (2008) conducted a case-study of online information-sharing amongst postgraduate
students, finding high levels of identity and trust within this small group. Wilson et al. (2006)
used experiments to test the role of trust in team-working within a work environment. However,
no literature was identified where a similar research environment had been created within the
transport field.
Ling et al (2005) used an existing online community (a US movie review forum) as their focus of
study, but set up experimental groups of members to study the factors which motivate users to
make contributions to such communities. This research was informed by social-psychological
Literature Review
theories, but in a long discussion of methods, the authors conclude that social science theories
are not always helpful in guiding the design of online community features, because such theories
do not allow for real-world complexity. For example, a design feature may influence multiple
psychological states and group processes, which each have multiple determinants, and this
cannot be adequately explained by any single social-psychological theory. Ling et al. (2005)
argue that, in their quest for abstract and generalisable theories of behaviour, socialpsychological theories are sparse, with typically only a small number of variables, which are
usually tested using the experimental method. This cannot adequately explain the full
complexity of human behaviour in context. This adds strength to the argument held in this thesis
for more in-depth, qualitative research which explores the complexity of social psychological
variables within a real-world context.
The only examples found in the conducting of this review where technology was used to set up
an experimental, online space for a qualitative study of social interactions within the context of
travel were within technology studies (e.g. human-computer interaction), where prototype
technologies are often „trialled‟ to see what happens when people use them outside the confines
of the laboratory. These studies are sometimes referred to as "field trials" (e.g. Esbjörnsson et
al., 2004; Axup et al., 2005). Esbjörnsson et al. (2004) conducted qualitative research with
(motor)bikers in Stockholm, who were given a new device to communicate with one another for
a limited period. The design of the device was informed by an ethnographic study in the
Stockholm biker community. As well as technical matters, findings covered issues of identity,
community and enjoyment of the biking experience, which were reportedly enhanced through
use of the technology. This raised the question of whether similar social processes might be
observed within a web-based environment for the sharing of local, informal, travel information
(Phase 2 of the research for this thesis).
Chapter Summary and Research Gaps
These findings and identified research gaps imply that the role of informal, socially acquired
information in travel behaviour is an unexplored field within transport studies. Much research
within transport studies has examined the influence of formal types of travel information on
decision-making, but little is known about the role of informal information, nor the processes of
word-of-mouth information diffusion, in the context of everyday travel. Moreover, theory and
empirical research on information-use concentrates on its role within individual decision-making,
neglecting both social factors and the role which information might play prior to the initiation of
active decision-making. Whilst social-psychological theories have been widely applied to the
study of travel behaviour, no studies were identified in this review where theories were
Literature Review
specifically applied to the social transfer of travel information. Moreover, the use of socialpsychological theory within travel behaviour has predominantly relied on quantitative methods,
typically seeking to measure pre-defined variables in order to test the applicability of theories.
Research in this field has less frequently sought a more in-depth understanding of socialpsychological constructs within the complexity of „real lives‟. Some studies of online communities
within this literature review have also identified weaknesses in the quantitative approach to
studying social-psychological mechanisms within an online environment.
The literature review has demonstrated that there are numerous behavioural theories within
social psychology and other disciplines which, together, offer a conceptual framework within
which informal travel information and word-of-mouth processes might be better understood. The
application of such theories and constructs – using qualitative methods - within this thesis
represents a novel approach, both theoretically and methodologically. A wide range of socialpsychological theories was presented in Section 2.2, many of which contain constructs relevant
to the field of study, but no single theory stood out clearly as one which should be applied above
all others. Moreover, similar constructs appear in different theories, albeit with slightly different
definitions (e.g. social and subjective norms; self-concept and self-categorisation). Therefore it
was decided to take an open approach in this thesis by seeking to identify a range of constructs
within the data, rather than „testing‟ a specific theory or group of theories.
Subjective information received by word-of-mouth (including electronic sources) has been shown
to be highly influential in people‟s choices about leisure travel, but has received little attention so
far in transport studies with regard to more everyday travel behaviour. This may reflect the
traditional dominance of instrumental-reasoned explanations for everyday transport choices, and
the assumption that formal sources and types of information are most appropriate for facilitating
these utilitarian choices. In contrast, holiday travel behaviour may be less affected by the
practical constraints of everyday life and more strongly influenced by the subjective information
and advice provided by other people - for example, intrinsic enjoyment of the travel experience
may be a greater consideration than simply getting to a destination quickly. It may also be harder
to access formal information about travelling in a distant and unfamiliar place, so word-of-mouth
becomes a relatively more important information source. It might therefore be conjectured that,
compared with leisure travel, everyday utilitarian travel behaviour might be more strongly
affected by practical constraints, habit, and past (personal) experience, and perhaps less so by
the types of social and subjective information which tend to be transmitted through word-ofmouth. However, little is known about the latter area; a gap which this thesis seeks to fill.
As informal information and word-of-mouth diffusion represent unexplored areas within transport
studies, a flexible research design and a qualitative research methodology were chosen for the
Literature Review
research in this thesis. The overall research approach and the specific methodology selected for
the exploratory phase of empirical research are explained in the following two chapters.
The research approach in this thesis
Chapter 3 : The research approach in this thesis
This chapter sets out the overall research design, the philosophical traditions which informed the
ontological and epistemological approaches, and the reasons for a two-phase approach to the
empirical research.
Research design
The exploratory nature of this research area led to the selection of a flexible research design. A
flexible design evolves as the research proceeds, and data are typically qualitative. A „good‟
flexible design requires rigorous data collection procedures, analysis and report-writing, framed
within assumptions of an evolving design, the presentation of multiple realities, and a focus on
participants‟ views (Robson, 2002). A flexible design was considered to be suitable for this
research because it addresses a relatively unexplored area where „bottom-up‟ theory building,
based on rich qualitative data, was first required before more detailed research parameters
could be defined. The flexible research design is consistent with the researcher‟s philosophical
stance of subtle realism (Hammersley,1992), which accepts the existence of an external reality
independent of individual subjective understanding, but suggests that this reality is only
“knowable” though subjective interpretation of both researcher and research subject (Snape and
Spencer, 2003). Much travel behaviour research is deductive in nature: the researcher starts
with a behavioural theory and uses this to develop and test hypotheses (Clifton and Handy,
2001). This is particularly the case where theories from social psychology have been used to
explain particular travel behaviours (e.g. Bamberg and Schmidt, 2003; van Vugt et al.1995, van
Vugt, 1997, van Lange et al. 1998). However, a more inductive strategy was adopted for the
present research. Although the research design was informed by a number of constructs and
theories from social psychology, an open, exploratory approach was adopted with the aim of
developing theory through analysis. Within travel behaviour research, Clifton and Handy (2003)
suggest that “While deductive research can involve quantitative techniques, qualitative
techniques, or both, inductive research generally relies on qualitative approaches” (p293).
A qualitative approach was chosen for the first phase of this research in order to seek, through
the explanations of research participants, a better understanding of the social context of travel
information and some of the social and psychological processes which might be involved. The
aim was to identify promising themes, travel contexts and behavioural constructs for further
study, rather than to test a particular theory or correlate information inputs with specific
behavioural outcomes. As Banister et al. (1994) note:
The research approach in this thesis
“Qualitative research is part of a debate, not fixed truth….an exploration, elaboration and
systematisation of the significance of an identified phenomenon” (p.3).
There is some novelty in this approach, as many (quantitative) studies of behavioural issues in
transport make use of a set of behavioural assumptions made by the modeller (based on
theory), but less research has been undertaken within the discipline to provide a deeper
understanding of these assumptions. Clifton and Handy (2003) identify an important role for
qualitative research in improving understanding of the complexities within travel behaviour:
“Yet the more we understand about travel behaviour, the more we recognise how much
there is that we do not understand. As one question is answered, new questions
emerge, and our appreciation of the complexity of travel behaviour grows. Qualitative
methods, used in conjunction with quantitative approaches or on their own (…), offer a
powerful tool for helping us understand these complexities” (p.283).
Before embarking on the detailed research methodology, consideration is given in the next
section to the epistemological and ontological foundations of this thesis, drawing principally on
the philosophical traditions of qualitative psychology.
Qualitative research and transport: some philosophical issues
Philosophical foundations
Given that the literature review has demonstrated a lack of any strong tradition of qualitative
research within transport studies, it was clear that the philosophical traditions of other disciplines
should be explored in order to inform and establish the philosophical grounding, and resulting
approach to data analysis, of the current research. For example, Clifton and Handy (2003) write
that “although qualitative methods have been employed in transport research, the field has been
predominantly entrenched in the quantitative paradigm for some time. This is not surprising
given the historical dominance of science and engineering in shaping the field of transport
studies” (p.298). Eight years later, Freudendal-Pedersen et al. (2010) state even more strongly
that "there is no qualitative research tradition in transport research" (p.25), arguing that transport
research has focused on physical mobility, and therefore on technology and modelling. Locating
themselves within the mobilities paradigm (e.g. Urry, 2000, 2002, 2004) these authors, along
with others in the mobilities field, elect to apply methods and knowledge from classical sociology
to transport research. However, as much of the theoretical grounding of this thesis has been
provided by social psychology, ontological and epistemological considerations have been
principally informed by the philosophical traditions - and corresponding data analysis methods 42
The research approach in this thesis
of qualitative psychology (whilst also acknowledging that many of the intellectual trends which
have shaped psychology have been influential across the social sciences). Even in psychology,
however, it should be noted that qualitative research still runs counter to the mainstream, rooted
as this discipline is within the experimental method. Writing as recently as 2008, Giorgi and
Giorgi remark that "psychology is extremely conservative in its interpretation of science, and one
departs from conventional criteria at great risk" (p.27).
The following section outlines some of the key developments in qualitative psychology over the
last century in order to provide a context for the identification of an appropriate ontological and
epistemological approach for this thesis, and for the data analysis in particular.
The development of qualitative psychology: from phenomenology to the discursive turn
Ashworth (2008) identifies three overarching schools of thought about the conceptualisation of
the subject matter of qualitative research in psychology. Firstly, for some, qualitative research
should aim at discovering the objective variables involved in the human situation; this has
parallels with the „strong‟ version of a realist ontology, which claims that there is an external
reality which exists independently of people‟s beliefs or understanding about it (Snape and
Spencer, 2003). Secondly, a person's understanding of the world may be conceptualised as a
set of quasi-linguistic propositions (or social interpretations); hence the object of study should be
the social nature of the constructions of the world which guide people's thoughts and actions.
Thirdly, qualitative psychology may conceptualise the person‟s grasp of their world in terms of
perceptions or meanings, which may be either socially shared or idiosyncratic. This
understanding of the focus of qualitative psychology is a cornerstone of phenomenology, in
which researchers try to describe and understand an individual‟s experience within their own
lifeworld. The latter two approaches may be equated with different strands of the idealist
position, which holds that reality is only knowable through the human mind and through socially
constructed meanings (Snape and Spencer, 2003). Of the three viewpoints, it is the third, with its
emphasis on meaning, which has influenced the ontology within this thesis most strongly,
although a phenomenological position per se has not been adopted, as will be explained in
Section 3.2.2.
Founded by Edmund Husserl (1859-1938), the principal philosophical basis of the
phenomenological movement was a rejection of the assumption that the focus of study should
be that which lies behind, or is more fundamental than experience. On the contrary, Husserl
maintained, investigation should always begin with what is experienced. Phenomenology does
The research approach in this thesis
not deny the existence of an „underlying reality‟, but employs the methodological tactic of
"bracketing" the question of reality separate from experience. The primary object of study is
considered to be the thing as it appears, that is the „phenomenon‟. Therefore it does not
contradict a realist ontology, but shifts the emphasis of study to the ways in which reality is
experienced. Husserl understood experience as a system of interrelated meanings (a Gestalt),
the totality of which are described as "the lifeworld" (Husserl, 1936/1970, cited in Ashworth,
2008). "In a nutshell, phenomenology insists that the daffodils are indeed different for a
wandering poet than they are for a hard-pressed horticulturalist." (Ashworth, 2008, p.12).
Equally, in the world of transport, a stretch of coastal railway is a different phenomenon for the
daydreaming leisure traveller than it is for the railway engineer.
Idiographic Psychology
A further influence on the development of qualitative psychology was the work of Gordon Willard
Allport (1897-1967). Allport was concerned that psychology should not neglect the unique in
individual experience and behaviour; the idiographic, focusing on the interplay of factors which
may be quite specific to the individual, is as important as the nomothetic – general dimensions
on which individuals vary (Allport,1962, cited in Ashworth, 2008). Allport did not restrict his
interest to the qualitative approach alone, however, recommending that as many different
methods as possible should be used to study an individual person.
The World as Construction
Phenomenology and idiography both lead us to view the person as a perceiver. Perception
provides direct access to what is being experienced; hence perception is not a construction or
representation. Although people are not necessarily passive perceivers - they pay attention
selectively, make choices, and their perceptions are always coloured with meanings which relate
to their life worlds – the phenomenological and idiographic approaches do not lead us to regard
the person as constructing their lifeworld. Social constructionism was an alternative school of
thought which gained prominence in the middle of the 20 century, becoming associated with
approaches such as symbolic interactionism, hermeneutics and postmodernism.
Symbolic Interactionism
The work of George Herbert Mead is regarded as a cornerstone of the constructionist orientation
in qualitative psychology (Mead, 1934, cited in Ashworth, 2008). During the 1950s and 1960s
this was absorbed into the school of social research known as symbolic interactionism, a highly
„social‟ approach which postulated that mind and self are both products of social interaction. An
The research approach in this thesis
infant‟s capacity for thought and sense of self only develops through a relationship of
communication with caregivers; therefore, thought arises in a social process. Only later can a
person‟s thoughts and self-concept become individualized, and this is, in any case, largely
dependent on the use of the social tool of language. Importantly, language and other systems of
symbols are of central concern to Mead, and linguistic symbols are a system of socially shared,
not idiosyncratic, meanings. Social interaction precedes thinking and selfhood, which are both
built with social materials. Once a child has acquired the ability to reflect on his or her own
action, he or she is able to build a self-concept or identity. Having developed the capacity for
mind and self as a result of interaction, the individual is then able to develop selfhood and
personal tendencies of thought in a relatively autonomous way within the social context in which
he or she is located.
Symbolic interactionism contends that the capacity for self-reflection develops through the child's
observation of other people's reactions to his or her behaviour. These ideas were influential in
the development of Albert Bandura‟s Social Learning Theory, which maintains that a child learns
by modelling the behaviour of others, and this occurs through the medium of symbols,
particularly through verbal representation systems; therefore, behaviour is learned symbolically
through the central processing of response information before it is performed (Bandura, 1977).
The symbolic interactionists contributed an extremely social outlook to qualitative psychology,
arguing that the person is first of all a member of society, and only later becomes an individual.
This leads to the argument that it is appropriate for qualitative psychology to explore the
symbolic systems of society, not just linguistic systems, but also those which are embedded in
the practices of the culture. This approach would lead one to conceptualise the car, for example,
as not just a means of transport, but a cultural artefact ripe with symbolic meaning (e.g. freedom,
independence and social status). Therefore, Mead‟s approach was an important building block
for discourse analysis and discursive psychology. Qualitative methodology (often ethnographic,
including participant observation) is often seen as the most appropriate approach for research
arising from symbolic interactionism and related areas. Symbolic interaction also influenced the
development of grounded theory and its specific framework for analysis (Strauss and Corbin,
Interpretation Theory: Hermeneutics
Having addressed the question of how the subjects of research are constructing their lifeworld,
constructionist thinking then asks: "what processes of construction have the researchers
themselves engaged in, in order to come up with the findings they have presented?". This is the
area of hermeneutics (which may be described as a theory of interpretation). In the
The research approach in this thesis
constructionist school, all science is a matter of construction; and, therefore, the conclusions of
research activity must be regarded as interpretations. We live in an interpreted world, and we
ourselves are interpreters (Ashworth, 2008). By stressing the active role of the researcher in a
dynamic process of knowledge generation, the hermeneutics approach offered a new view of the
meaning of data in qualitative research, and one which was influential in, for example, the
development of interpretive phenomenological analysis (IPA). As practitioners of IPA, Smith and
Osborn (2008) refer to a two-stage interpretation process, or a double hermeneutic: "The
participants are trying to make sense of their world; the researcher is trying to make sense of the
participants trying to make sense of their world” (p.53). This understanding of empirical
qualitative data was one which was influential within this thesis.
The Discursive Turn and Postmodernism
In the first half of the 20th century, a turn in Western philosophy led many thinkers to begin to
see language use as "ontologically primary". The premise was that language does not simply
reflect the world of experience, but rather our world is constituted by our shared language. Thus,
language becomes, in a sense, the prime reality, and should therefore be the focus of study.
This philosophical stance has influenced qualitative methods such as narrative analysis,
conversation analysis and discourse analysis. This radical form of social constructionism
suggests that the "perceptual tendency" - the idea that we can directly describe experience should be abandoned, as experiences cannot but be shaped by constructions of events. Thus,
qualitative psychology cannot reveal a lifeworld, but only a network of elements, which can only
derive meaning from their position within the total system. Hence, all is construction, or "all is
text" (Derrida, 1974, cited in Ashworth, 2008). This is a central tenet of postmodernism.
Postmodernism rejects the modernist assumption that our perceptions and constructions relate
to the real world, and that ongoing progress is possible in research. The fundamental tenet of
modernity is that there is a non-negotiable, solid truth or reality about which ever more accurate
knowledge can be acquired. Thus, the researcher is engaged in a process of elaborating the
structure of scientific constructs in a way which comes closer and closer to the „truth‟ of actual
reality. The knowledge advances incrementally, by following the recognized criteria of scientific
research and scholarly activity. This is the paradigm within which natural science and technology
is located, as well as the social and political world. Transport research, whether addressing the
technological, the behavioural or the political, is solidly located within the modernist paradigm.
For postmodernists, these criteria of validity no longer exist, as the connection between reality
and human constructions has been abandoned. The idea of progress no longer applies, as
there is no objective standard against which to evaluate an innovative theory, practice, product
The research approach in this thesis
or policy. Most psychology is modernist in its approach, but postmodernism questions this view,
suggesting that psychology can no longer regard itself as lying "outside human society, looking
in". It is just one of many discourses within a culture in which it resides. A consequence for the
researcher is that he/she must see him or herself as part of this web of cultural construction. The
researcher cannot be detached from the culture and society in which he or she is situated, so
research becomes a joint product of both researcher and researched. In a weaker form, this
view is connected to the broader concept of „reflexivity‟ within research. Although a
constructionist position is not adopted in this thesis, reflexivity as a principle is strongly
This section has outlined some of the key philosophical trends shaping qualitative psychology
over the last century. The next section will attempt to show how some of these philosophical
traditions have influenced the approach within this thesis. The links between these traditions and
the chosen method of data analysis will be discussed in the next chapter.
Ontology and epistemology in this thesis
It may have become apparent from brief observations in the sections above about the
relationship of the present thesis to these specific philosophical approaches, that an affiliation
has not been made to the „strong‟ versions of either the perceptual (e.g. phenomenological) or
constructionist (e.g. postmodernist) tendency within qualitative psychology, as neither of these
traditions are considered by the researcher to be exclusively appropriate for the current study.
As this thesis spans a number of academic disciplines, it is believed instead that there are
advantages to drawing on a variety of traditions, which in any case have many overlaps, by
adopting a pragmatic and mixed approach.
Ashworth (2008) supports a pluralist approach, holding that the different philosophical
tendencies “… have a number of emphases, but it is arguable that the richness of the human
condition is such that no one tendency would encompass the whole (…..). Pluralism in
qualitative psychology is to be valued." (p.25). Many researchers argue for a pragmatic
approach to social research in order to ensure that it is relevant to contemporary society without
becoming too deeply enmeshed in very pure philosophical traditions. For example, Bent
Flyvbjerg (2001) argues for a phronetic social science based on practical judgement and
common sense, suggesting that this may contribute most effectively to social and political
development. Snape and Spencer (2003) argue for what they label a "toolkit" approach to social
research, which focuses more on finding the appropriate method for addressing specific
research questions, rather than being too concerned with (or constrained by) the underlying
philosophical debates.
The research approach in this thesis
Snape and Spencer and (2003) point to an important contextual factor which may help to explain
why much research in the social sciences adopts a practical approach to epistemology. This is
that most social research is commissioned and funded by public bodies who wish to use the
findings to inform policy and practice. The dominant research paradigms within this context
have tended to be quantitative, and this is undeniably the case within transport research. This
means that qualitative research in the social sciences has been strongly influenced by some of
the expectations and practices which have grown out of quantitative research: "evidence" must
be rigorously collected and analysed, be valid, able to support wider inference, must be as
neutral as possible, and the ways in which interpretations have been reached must be clearly
defensible. Snape and Spencer (2003) suggest that they, and many other qualitative
researchers, do not fit neatly into any one „school‟ of qualitative research, and instead borrow
from different traditions within the social research field generally. They do, however, warn
against the temptation to adopt certain practices which are generally acknowledged, but then
failing to discuss explicitly the beliefs underlying them. This can make it difficult for others to
judge the quality of their research. Moreover, there may be tensions if researchers attempt to fit
the „square peg‟ of quantitative terms such as evidence, inference and generalisability into the
„round hole‟ of a more interpretative epsitemology (and this may be one reason why many
research reports appear to „gloss over‟ the underlying beliefs of the researcher).
The ontology underlying this thesis might be described as a version of realism with a strong
idealist flavour. This corresponds to subtle realism (Hammersley, 1992), which accepts the
existence of an external reality independent of individual subjective understanding, but suggests
that this reality is only „knowable‟ though the subjective interpretation of both researcher and
research subject. Thus, the data were understood to represent a combination of behaviours,
attitudes and understandings as expressed by the participants in a research setting, i.e. the
participants‟ subjective interpretation of an external reality. The status of the data was therefore
one of representing social phenomena which exist apart from the setting in which the data were
collected (i.e. the interview, focus group or website). Therefore, „reality‟ has not been bracketed
off, nor has the object of study been limited to the lifeworld of participants, as a
phenomenological approach would imply (although understanding the meaning which
participants ascribe to their experiences is indeed a vital aspect of this thesis).
The strong version of constructionism, where there is deemed to be no single underlying reality,
has also been rejected. This partly reflects the tradition of transport research: this is an applied
field, benefitting from public funding, and orientated towards informing policy and practice. In this
The research approach in this thesis
context, a denial of an objective, external reality would seem incongruous – an intellectual nicety
indulged in by academics and divorced from the everyday transport concerns of real people. Yet
despite this, a doctoral thesis should represent a rigorous intellectual exercise; although it may
have (and, arguably, should have in transport) „real world application‟, it is not an applied
consultancy project. It is for this reason that an attempt has been made to clarify the ontological
and epistemological and assumptions which shaped the design, execution and conclusions of
the research.
The epistemological approach within this thesis falls within the broad category of interpretivism.
In opposition to positivism, interpretivism asserts that facts are not separate from values, and
research findings are always influenced by the perspectives of the researcher. However, by
being „reflexive‟, the researcher can attempt to identify and acknowledge his or her own
assumptions and values. Because the social world is not governed by law-like regularities, but is
mediated through meaning and human agency, the social researcher seeks to understand the
social world using both his/her own interpretations, and those of the participants (Snape and
Spencer, 2003). In this sense, the epistemological stance encompasses some aspects of
phenomenology (exploring participants‟ meanings), idiography (exploring the unique in human
experience and behaviour), and hermeneutics (interpretation).
Consistent with the chosen epistemological approach, the initial reading of the data was
principally an interpretive one, in which the researcher explores the participants‟ interpretation or
understanding of social phenomena and their own behaviours, attitudes and motivations, as well
as constructing her own interpretations of their accounts (Mason, 2002). Hence, the approach in
this thesis concurs with Kozinets‟ (2010) critique of the term “data collection”, which “seem(s) to
imply that these things “data” are scattered about, like leaves on the ground or documents on a
table, and that your job is simply to gather them up and “collect” them.” (p.95). The reading of
the data was reflexive "a reflexive reading will locate you as part of the data that you have
generated" (Mason, 2002, p.149). In line with the double hermeneutic previously described
(Smith and Osborn, 2008), the analytic account could be described as the researcher's
interpretation of participants‟ interpretations. The researcher was not simply “gathering”
interpretations and understandings which already existed in the participants‟ minds, ready to be
retrieved in response to neutral questioning, but also encouraging them to think about their own
and other people‟s travel behaviour in ways which they had not necessarily done before.
Having given detailed consideration to philosophical issues underpinning the research, we turn
in the final part of this chapter to the reasoning for separating the PhD into two research phases.
The research approach in this thesis
Rationale for two research phases
In the absence of previous research in this area, there was a need for initial, exploratory
research about the use and effects of informal types of information and word-of-mouth delivery
processes, in order to understand the general context in which it is used (what, when and by
whom?), and to start to identify perceived effects and underlying social mechanisms which might
explain these effects (research questions 1, 2 and 3). Only then would it be possible to move the
focus of study to a specific applied area.
Following the exploratory research using interviews and focus groups, the next task would be to
study how far the findings might be validated in the context of a real-world traveller information
system, as well as exploring some of the initial findings - especially social-psychological
mechanisms - in greater depth (research question 4). Recognising that interview and focus
group methods can have limitations, Phase 2 would include observation of the actual use of an
information system, as well as reported behaviour, thereby offering opportunities for
methodological triangulation. Conclusions might then be drawn from both phases of the research
which could be of relevance to advanced traveller information systems more widely (research
question 5) – although further research would be required (outside the remit of this thesis) to test
the wider validity of the findings in different contexts. Although the main research questions had
been defined at the start of the research, the movement from the first to the second phase
constituted an iterative process whereby key areas of interest emerging from the exploratory
research were formulated into detailed research questions for further exploration within the casestudy. This process will be described in greater detail in Section 6.1.2.
Chapter Summary
This chapter has outlined the overall research approach in the thesis: a flexible design
employing predominantly qualitative methods, based on an ontology of subtle realism and an
interpretative epistemology. The lack of a strong qualitative tradition in transport studies, coupled
with the influence of social psychology theory within this thesis, led to the identification of
qualitative psychology as the field which would provide the main philosophical grounding for the
research. Having outlined the historical development of some key philosophical tendencies in
qualitative psychology, we have explained that, rather than drawing on one specific tradition, a
pluralist approach (drawing on aspects of phenomenology, idiography and hermeneutics) was
selected for the current study. Finally, it was explained that the unexplored nature of the
research topic necessitated a two-stage approach, whereby exploratory research would first
seek to uncover promising themes which might then be studied in greater detail within an
applied context.
Empirical Research, Phase 1: Exploratory
Interviews and Focus Groups
Methodology, Phase 1
Chapter 4 : Methodology, Phase 1
The first empirical research phase set out to explore the first three research questions through a
qualitative study of the social context in which people acquire and offer informal travel
information through word-of-mouth. It explored the role of informal information in the formulation
of beliefs, attitudes and intentions and their influence on travel choices, drawing on a number of
constructs and models from the field of social psychology and decision theory. Unlike many
studies within travel behaviour research, which apply theories to explain observed behaviour and
identify the antecedent factors, the aim of this phase of the research was to explore, using an
interpretative epistemology, people‟s own detailed explanations of some of the social and
psychological processes which contributed to their travel behaviours.
A qualitative methodology was therefore designed to address Research Questions 1 to 3:
a) Context: the social transfer of travel information:
1. From whom, and in what circumstances, do people acquire travel information, and with
whom do they share it? What types of information are conveyed through word-ofmouth?
b) Mechanisms: the influence of word-of-mouth information in travel behaviour:
1. What is the perceived role of word-of-mouth information in the travel decision process?
How do informal and formal types of information interact?
2. How do particular social-psychological factors (e.g. social norms, social identity, prosocial values, trust) appear to influence the use and effects of word-of-mouth
Selection of methods
The chosen methods of generating data for this phase of the research were in-depth interviews
and focus groups. These were selected as a means of drawing out themes in a relatively
grounded manner from participants‟ explanations of their own behaviour and motivations, and
their beliefs about the motivations of others. Both interviews and focus groups were undertaken,
as each method can generate a different form of data. Within a one-to-one interview, material is
generated through the interaction between researcher and interviewee; a semi-structured
interview allows key topics to be covered with the flexibility to ask follow-up questions which
probe an individual‟s underlying reasons, feelings, opinions and beliefs. In a group environment,
Methodology, Phase 1
data are generated through interaction between participants, which is particularly pertinent to a
study of word-of-mouth information-sharing. Whilst a focus group offers less opportunity to probe
individual views, greater spontaneity may arise from the stronger social context (Ritchie and
Lewis, 2003, Kruegar, 1994).
The role of the researcher was given some thought during the design as well as the conducting
of the interviews, seeking consistency with the researcher‟s ontological and epistemological
stance, as described in the previous chapter. A decision was taken not to strive for complete
objectivity or neutrality in interactions with the interviewees before, during or after the interviews,
but rather to acknowledge that the nature of the interactions would have some effect on the data,
and to attempt to use this in a positive way which might contribute to the richness of the data.
This decision was influenced by Oakley‟s (1981) critique of the traditional view of interviewing
as a mechanical instrument of data-collection, or a specialised form of conversation in which one
person asks the questions and another gives the answers. Oakley elaborates on the importance
of the one-to-one relationship between interviewer and interviewee, the personal meanings
inherent in each social interaction, and the advantages in terms of rapport of responding to
interviewees‟ questions within the interview. In line with the hermeneutic approach outlined in the
previous chapter, data were being generated and not simply collected. The more positivist role
of the interviewer as an impassive and neutral questioner, refraining from betraying any
response to an interviewee‟s answers (e.g. by nodding the head or smiling), not giving away any
personal information, and not answering interviewees‟ questions, was therefore rejected. The
aim was to use some of these conversational practices in a controlled way in circumstances
where they were felt to put the interviewee at ease and encourage fuller responses. This was felt
to be all the more appropriate because the researcher was a member of the same community
(i.e. the university) as the participants, which meant that each was party to a degree of preexisting knowledge about the other, even if they were not personally acquainted. It also meant
that some participants believed the interviewer and other focus group participants to possess
information which might be helpful to them, and therefore occasionally asked questions during
the interviews and focus groups.
As well as the philosophical reasoning, this approach was also felt to offer practical advantages.
Participants were given some information about the research topic and two general areas of
questioning prior to the interview or focus group. This was largely because the pilot interviews
had shown that some people found it difficult to think of examples of relevant experiences when
„put on the spot‟ during the interview, which made in-depth follow-up questioning more difficult.
This could be interpreted as a finding in itself, suggesting that people may not pay much
attention to everyday interactions about travel. The interviewees were therefore invited to think
Methodology, Phase 1
about certain trip examples beforehand, and these examples could then be probed during the
interview to reveal more depth of thought. A possible drawback of this approach is that some
answers were less spontaneous than they would otherwise have been.
As alluded to above, the interviews and focus groups were not only a forum in which participants
offered their own reflections on the process of informal information-sharing, but also a means
whereby they could share travel information with one another (and to a lesser degree with the
researcher). The focus groups in particular were an opportunity for participants not only to
discuss their experiences of information-sharing, but also to engage in the very process they
were discussing. This meant that data took the form not only of behaviours, attitudes and
understandings as expressed by the participants, but also, to a limited extent, their direct
observable behaviour in engaging in the processes being studied. As the principal
methodological aim of this phase of this part of the study was to understand participants‟
interpretations, rather than to observe their behaviour directly, the interview and focus group
transcripts were analysed from the former rather than the latter perspective. Observation of
actual information-sharing behaviour was undertaken in the second phase of the research (the
Population, sampling and recruitment
In-depth interviews and focus groups were conducted among new members of staff and
students at the University of the West of England, Bristol (UWE). Participants took part in either
an interview or a focus group, according to preference. Limiting the population to new
employees and students within a single organisation offered the opportunity to explore and
compare the different forms of information used by people travelling to a common destination for
the first time, as well as to consider social norms and social identities within a particular
organisation, or sub-groups within it. A university was also thought to provide fertile ground for
this topic, as research has shown a positive correlation between educational level and travel
information-use (Farag and Lyons, 2008b). With regard to word-of-mouth information, the 2007
UWE travel survey had revealed that 17% of those surveyed had obtained information about
travelling to the university by “asking someone they knew”. The university provides a useful
population from which to draw a sample because its membership is constantly in flux. Starting at
a university, either as a student or employee, represents a life change when people might be
susceptible to changing their previous travel behaviour, and are likely to seek out travel
information actively. Recruiting within the university also allowed a certain degree of
comparability across cases.
Methodology, Phase 1
Consistent with the qualitative approach taken, a judgemental or purposive sampling strategy
was employed (Blaikie, 2000). This is a form of non-probability sampling which does not seek full
representation of the target population; the aim was not to make generalisations about the whole
population, but rather to identify interesting themes to be carried forward for further study.
Employees were recruited at induction sessions and via an email list of new staff. Students were
recruited at induction sessions and at the Freshers‟ Fair. Participants were offered £10 as a
reward for taking part. The sample comprised employees who had taken up a post at the
university within the past six months, and students who had started a course there within the
previous month. The sample was selected in order to provide a diversity of socio-demographic
characteristics and travel behaviours, to allow a range of different perspectives to emerge. By
including a range of ages, gender, employees and students, and preferred modes of transport in
this sample (a simple form of quota sample), it was hoped that a significant sample bias would
be avoided. However, a certain degree of bias was accepted in drawing the sample from a
specific (i.e. university) population, which has certain social characteristics, for example in terms
of educational attainment, which make it unrepresentative of the wider population.
In total, thirteen interviews of approximately one hour, and two focus groups of 90 minutes were
undertaken. Ages ranged from 19 to 60 (mean 34) and gender was balanced for the interviews
but skewed towards women in the focus groups. One focus group comprised only employees,
and the other only students, as social heterogeneity may sometimes inhibit some participants
(Ritchie and Lewis, 2003). Age, gender and primary transport mode for commuting are shown
below in Table 4-1.
Table 4-1: Composition of Phase 1 sample by age, gender and primary transport mode for
commuting to the university
Primary transport mode for
Age range
cycle walk
Recruiting a balanced mix of participants was not straightforward. Among the employee group,
far fewer men volunteered than women, and the majority volunteered to take part in one-to-one
interviews rather than a focus group (hence the employee focus group comprised only women).
Data collection took place in two phases: the first in May 2008 (employees) and the second in
Methodology, Phase 1
October 2008 (new students). To redress the gender imbalance, male students were recruited
more actively than female students in the second round. To encourage involvement in a focus
group, the financial „reward‟ for focus group participation was increased from £10 to £20. Some
of the employed participants did not accept the reward, and others only did so when told that the
money was provided from a research budget rather than directly from the researcher. Whether
they were influenced by the increased financial reward, or were simply less comfortable with a
one-to-one interview, the majority of students volunteered to take part in a focus group. A sizable
number of the student volunteers dropped out between verbally agreeing to take part and setting
up firm arrangements. One of the confirmed (male) participants failed to show up for the focus
group, and another had to withdraw at the last minute due to a timetable change, but agreed to
take part in an interview instead. However, despite these set-backs, it was felt that an adequate
diversity of socio-demographic characteristics and travel behaviours had been achieved.
Content of the interviews and focus groups
Four pilot interviews had been undertaken beforehand and had revealed that some people found
it difficult to remember the travel information sources they had used, particularly word-of-mouth.
Therefore, all participants were emailed two days before the interview or focus group, inviting
them to reflect on any sources of information they had used for their first trip to UWE and for
another unfamiliar trip they had planned during the previous six months. They were also emailed
a short summary of the project and a consent form two days beforehand, in order to save time
when they arrived.
Topic guides were developed for both the interviews and focus groups, based on the research
questions and relevant literature, both theoretical and empirical (see Appendix C). Particular
attention was paid to ways of exploring social-psychological constructs, with reference to the
phrasing of questions in previous studies (predominantly surveys). Clifton and Handy (2003)
highlight the importance of carefully crafting the questions in qualitative travel behaviour
research, noting that “a relationship of trust needs to be established within a relatively short time
frame”. However, they also note that it is to the benefit of researchers in this field that “many of
the lifestyle and preference issues that are of interest to transportation researchers are not
exceedingly sensitive in nature” (p296). It goes without saying, however, that sensitivity should
be used when asking personal questions, even when they apply to topics which may not seem
sensitive in nature.
The topic guides for the interviews and focus groups are attached as Appendix A and Appendix
B. Questioning and discussion covered the following general areas:
Methodology, Phase 1
Use of, and preference for, different types and sources of travel information in different
situations, including first trips to UWE, and interactions between the two.
Probing about word-of-mouth sources: what they learned; whom they received it from;
whom they offered it to; their motivations for doing so; and its perceived impact (if any)
on travel behaviour.
Probing of social and psychological factors which might underlie these processes (e.g.
social and subjective norms, pro-social values, self-concept, social identity and trust).
At the start of the focus groups, following general introductions and an outline of participants‟
travel behaviour to and from the university, photographs of different travel information sources
(including people) were displayed in order to stimulate discussion. During the interviews, social
and psychological factors were explored within the context of one or more specific examples of
informal information-use provided by the participant earlier in the interview. In the focus groups,
participants were invited to discuss and order a series of „sort cards‟, each containing a
statement which might help to explain why they would (or would not) take word-of-mouth
information into consideration, or why they might offer informal information to other person (e.g.
“I trusted the other person”, “I had things in common with the other person”, “I wanted to help the
other person”). Blank cards were also provided so that participants could add their own
statements. Thus, participants in the first focus group added “knowledge and expertise of the
other person” as a reason for considering word-of-mouth information, and this was added to the
set in the second group. The sort cards proved to be particularly helpful in stimulating open
discussion in the (female) employee focus group, but less so in the student focus group which
was less homogenous and where some (older, male) participants were more obviously
dominant in the discussion. Each focus group was attended by one of the researcher‟s
colleagues acting as an observer. The observers made notes of some of the non-verbal
communication amongst participants which the researcher (moderator) was unable to record,
and afterwards provided feedback to the researcher.
After each set of two or three interviews, notes were made by the researcher, reflecting on
aspects such as the rapport with each interviewee, quality of the interactions (e.g. whether they
were more familiar and less „neutral‟ when there was perceived to be empathy between the two
people) and the extent of any interviewer bias (e.g. questions being phrased in a leading way).
Whilst some of the questioning could be regarded as leading, particularly when unplanned
questions were asked spontaneously in response to participants‟ accounts, interviewees
generally appeared to give considered responses rather than what they guessed to be the
socially acceptable answer, although admittedly this cannot be known for certain (Bonsall,
Methodology, Phase 1
2009). The interviews were conducted in an open and non-hierarchical manner (interviewer and
participants all working/studying at the university and having a similar level of educational
background), which reduced the likelihood of interviewees‟ responses being greatly biased by
their perceptions of the interviewer. However, these issues were given due consideration and a
reflexive approach adopted when the transcripts were analysed.
The proceedings were transcribed verbatim by the researcher in two phases, using voice
recognition software, following each round of interviews and focus groups, and imported into
QSR NVivo – a computer aided qualitative data analysis package.
Data analysis
As indicated in Chapter 3, some thought was given to the ontological and epistemological
approach to be adopted, before the analysis was embarked upon. There are two broad camps
within qualitative analysis: firstly, those methods which are closely allied to a particular
theoretical or epistemological position, and secondly, those which are essentially independent of
epistemology and can therefore be applied across a range of theoretical approaches (Braun and
Clarke, 2006). Some methods belonging to the first category, such as phenomenological
analysis, interpretive phenomenological analysis (IPA), discourse analysis, conversation
analysis, narrative analysis and grounded theory (although this is not, strictly speaking a
„method‟, but rather a „research framework‟) have already been alluded to in Chapter 3. Just as
the philosophies underpinning these methods have contributed to the establishment of a
philosophical stance for this thesis, so too have the related methods influenced the thinking
about how the analysis of empirical data should be undertaken. Some specific analysis methods
were ruled out because of their incompatibility with the philosophical approach adopted in the
thesis. For example, in discourse analysis and conversation analysis, the focus of study is
language use and the discourses communicated through social interaction, but no claims are
made about any kind of „objective reality”. These methods are strongly associated with social
constructionism, and hence were deemed inappropriate within the more realist conceptual
approach of this PhD. Three methods which have influenced the present analysis are
phenomenological analysis, IPA and grounded theory; the influences of these methods on the
present analysis are outlined in Appendix D. However, the main analysis method finally selected
belonged to the second of the categories mentioned at the beginning of this paragraph - those
methods which can be applied across a range of theoretical approaches.
Methodology, Phase 1
Cross-sectional thematic analysis
Rather than adopting a method tied to a particular philosophical tradition, the principal method
applied in this thesis is one which is frequently employed - but often unnamed - in qualitative
social research: thematic analysis (Braun and Clarke, 2006). This selected method also draws
strongly on the related cross-sectional “code-and retrieve” method within the “Framework”
approach developed by the National Centre for Social Research (Ritchie et al., 2003), and the
method of cross-sectional analysis set out by Jennifer Mason (2002). Cross-sectional/thematic
analysis was used exclusively for the analysis of Phase 1 data, but for Phase 2 was
complemented by non-cross-sectional analysis of the data elicited through a combination of
observation, questionnaire and interview methods from a small number of specific individuals.
This method, whereby individual cases are analysed in a holistic way, rather than through
comparison with other cases, is also referred to as contextual, case-study or holistic analysis
(Mason, 2002), or vertical analysis (Freudendal-Pedersen et al, 2010), and will be explained in
Chapter 7.
Braun and Clarke (2006) concede that the process of thematic analysis is not necessarily
unique, and can be found in many types of qualitative research. Indeed the steps outlined here
bear a strong resemblance to those followed within phenomenological analysis and IPA, as
described in Appendix D. A thematic analysis involves the following phases (Braun and Clarke,
Familiarising oneself with the data: transcribing data (if necessary); reading through the
data and noting down initial ideas through a process of synthesis or data reduction.
Generating initial codes (also known as an index): coding interesting features of the data
in a systematic fashion across the entire dataset; collating data relevant to each code.
Codes may be "data-driven", in which case themes will emerge bottom-up from the data,
or "theory-driven", when the researcher might be reading the data with specific
questions in mind.
Searching for themes: collating codes into potential themes; and gathering all data
relevant to each potential theme.
Reviewing themes: checking if the themes work in relation to the coded extracts (level
1), and the entire dataset (level 2); and generating a thematic map of the analysis. This
involves refining the candidate themes.
Defining and naming themes: ongoing analysis to refine the specifics of each theme and
the overall story the analysis tells, generating clear definitions and names for each
Methodology, Phase 1
Producing the report: writing the report is the final opportunity for analysis.
The first three steps above might be referred to as the data management phase (Spencer et al.,
2003). The function of the codes is to organise the retrieval of sections of text for the purpose of
cross-sectional analysis (Mason, 2002). For example, in analysing the data from Phase 1, the
codes were arranged in a hierarchy under headings and sub-headings which reflected the basic
structure of the topics discussed in the interviews and focus groups (e.g. use of travel
information, receiving travel information, giving information). Within this structure which was
largely theory-driven, „bottom-up‟ codes were also allowed to emerge from the data. Once the
process of coding text in the transcripts began, more codes were added, some collapsed into
others, and others refined, resulting in around 100 codes. Some groups of codes were moved up
or down the hierarchy as they began to emerge as more or less prominent. The hierarchy of tree
nodes for Phases 1 and 2 are attached as Appendix E and Appendix L respectively. Each
transcript was then coded sentence by sentence using the NVivo package. Each code was
described and memos attached as ideas about the themes emerged.
The fourth and fifth steps (reviewing and defining themes) equate to a second stage in the
analysis, which Ritchie et al. (2003) term “descriptive accounts”. This involved reading through
the text assigned to each code, making comparisons between cases and looking for common
themes, associations and contrasts. Themes and linkages were recorded as diagrams and
elaborated through the process of writing. This evolved into a process of seeking explanatory
accounts of the themes which were emerging from the data. Ritchie et al. (2003) distinguish
between two types of explanation at an analytic level: those which are based on the explicit
reasons that are given by participants themselves (for example, participants make travel
decisions by weighing up options and making up their own minds); and those arising from
implicit reasons which are inferred by the analyst (e.g. why participants may wish to present
themselves as individual, rational decision-makers). Braun and Clarke (2006) define the two
approaches, respectively, as semantic and latent analyses. Different types of evidence were
used to generate and support explicit versus implicit accounts. In the former case, the evidence
appeared overtly in the reasoning within the participants‟ accounts. In the latter case, the
researcher drew upon a number of the strategies proposed by Ritchie et al. (2003): searching for
a possible underlying logic in what people said; using common sense to search for explanations;
comparing findings with those in other studies; or relating findings to a theoretical framework. In
reporting on the findings, attempts were made to link them to relevant theories and empirical
studies in the literature review.
Methodology, Phase 1
Generalisability, reliability and validity
The issue of generalisability of qualitative research - whether findings can support wider
inference beyond the sample or population of study - is an important one, but one where there is
very little consensus. Hammersley (1992) distinguishes between empirical and theoretical
generalisation. The former concerns the application of findings to other people or contexts, and
is also referred to as external validity (Robson, 2002). Theoretical generalisability involves the
building of theories from the specific study which may be more widely, even universally
applicable. Lewis and Ritchie (2003) categorise the development of policy recommendation from
the findings of a specific sample as an example of theoretical generalisation. They further
separate empirical generalisability into representational generalisability – whether findings from
the research sample can be generalised to the parent population; and inferential generalisability
– whether findings can inferred to other populations.
In the present research, the relative consistency of the findings at a higher level of abstraction
(such as participants‟ construction of themselves as rational decision-makers) may imply a
degree of representational generalisability from the sample to the parent population – i.e. the
University - although this cannot be „proven‟ in the same way as would have been achieved
through probability sampling. More important was the process of theoretical generalisation from
the present research through the elaboration of behavioural theory about (travel) informationuse. However, this understanding of theoretical generalisation differs from the positivist
conception of the formulation of universal laws, unrestricted in time and space. Rather it is
aligned to Lewis and Ritchie‟s view that:
“qualitative research studies can contribute to social theories where they have
something to tell us about the underlying social processes and structures that form part
of the context of, and the explanation for individual behaviours or beliefs. (…) The
degree to which the data from a study support existing theories can be assessed, by
comparing how well different cases “fit” within an established theory and how far it is
able to explain behaviour in individual cases. Those theories can then be developed and
refined so that they accommodate any newly found variations in behaviour or
circumstances identified through the research.” (2003, p.267)
A degree of inferential generalisation was sought in Phase 2 by extrapolating some of the casestudy findings to other possible information system settings. Such extrapolations are described
by Patton (2002, cited in Ritchie and Lewis, 2003) as:
Methodology, Phase 1
“…modest speculations on the likely applicability of the findings to other situations under
similar, but not identical conditions. Extrapolations are logical, thoughtful and problemoriented rather than statistical or probabilistic.” (2002, p584)
Reliability, or the replicability of the research findings if they were repeated in another study, is a
concept with which qualitative researchers are often uncomfortable, to the extent that a number
of alternative terms have been introduced. For example, Glaser and Strauss (1967) write of
trustworthiness, and Hammersley (1992) of consistency. Like reliability, validity is a concept
originating in positivism which is substituted by some qualitative researchers with terms such as
credibility and plausibility (Glaser and Strauss, 1967). Internal validity relates to the robustness
of the research design, or precision of a research reading, whereas external validity concerns
the extent to which results can be generalised to a wider population (Robson, 2002; Lewis and
Ritchie, 2003). The present research has attempted to address the threats to validity which can
emerge as a result of factors such as reactivity, participant biases and researcher biases
(Robson, 2002). This has involved strategies such as leaving an „audit trail‟ to ensure openness
and transparency, reflexivity (constant reflection on the research process and open reporting),
and triangulation both through comparison with theory and other empirical studies, and through
the application of different research methods within the project.
Chapter Summary
This chapter has described the methodology for the first empirical research phase. Interviews
and focus groups were selected as the research method. Twenty four people (15 women, 9
men) were recruited purposively from a population of new employees and students at the
University of the West of England. Thirteen people took part in a one-to-one interview, and two
90-minute focus groups were held: one comprising 6 employees and the other comprising 5
students. The interviews explored participants‟ use of informal travel information and word-ofmouth when they first started their work or study at the university. Examples of giving and
receiving information through word-of-mouth in other contexts were also explored, seeking to
provide an understanding of the influence of such information on beliefs, attitudes and travel
behaviour, and to identify social and psychological mechanisms through which this influence
might have occurred. Interviews and focus groups were transcribed verbatim and analysed using
a cross-sectional thematic analysis. The findings are discussed in the next chapter.
Findings, Phase 1
Chapter 5 : Findings, Phase 1
The main findings from the first research phase are summarised below under headings
corresponding to the first three research questions. This is followed by a brief discussion of
some of the other areas of findings which were deemed to contribute additional understanding to
the context of informal information-use. The exploratory phase provided context and brought to
light themes which would be explored in greater detail in Phase 2.
The social context of word-of-mouth information-sharing: who, what and when?
Research Question 1
From whom, and in what circumstances, do people acquire travel information, and with whom do
they share it? What type of information is conveyed through word-of-mouth?
The findings reported in the first two sections below - people and content - were answered
largely through participants‟ explicit (literal) accounts (explained in Section 4.3.1), whereas the
third part - circumstances - required a more interpretive analysis of implicit as well as explicit
The most frequent examples of the social contacts from whom travel information was acquired,
or to whom it was offered, prior to or after a trip taking place were: family members, friends, work
colleagues, course-mates, house-mates, neighbours, and fellow members of social clubs or
groups. Information was shared with people who might be categorised as both „strong‟ and
„weak‟ ties (Granovetter, 1973) – the most important factor being that the information-provider
had access to local knowledge (although the influence of such information may have been
affected by the relationship between the provider and recipient of information, as will be
discussed in Section 5.3). The social closeness of the people cited as having provided
information (and to whom it had been offered) also varied according to trip context; for example
advice on travel to and at holiday destinations tended to have been provided by friends and
family, whereas smaller, everyday (and „lower risk‟) trip details might be provided by any type of
acquaintance. Unsurprisingly, if information was required during a trip, participants travelling
alone might ask a fellow passenger, selecting whom to approach on the basis of how friendly or
knowledgeable they thought they looked. Several (younger) participants reported that they would
also contact friends or family for advice via mobile phone during a trip, particularly for help with
way-finding. Informal types of information had been obtained both pre-trip and on-trip from
Findings, Phase 1
professionals, such as transport employees or travel agents. Occasionally, online reviews and
blogs were consulted for informal travel information, usually when participants were travelling
abroad, but occasionally via message boards within online communities relating to a particular
interest or hobby closer to home.
The people most frequently cited as giving or receiving travel information before/after or during a
trip are depicted in Figure 5-1 and Figure 5-2 respectively. During pre-tip planning, or in
discussion after a trip, information had been both given to, and received from the strong ties to
the left of Figure 5-1 (although not necessarily between the same two individuals or within the
same interaction), but had flowed only one way from the weak ties to the right. Clearly, advice
might be given as well as received via internet forums, but the participants in this sample had not
posted travel information themselves. The categories of people in the centre of Figure 5-1 could
include both strong and weak ties, and information had flowed both ways with these groups. It
should also be noted that the categories of people are not mutually exclusively; for example,
many of the people described by participants as „friends‟ were also, for example, neighbours,
colleagues and course-mates. Figure 5-2 shows that information had been shared with other
travellers whilst en route, but had flowed one way from passers-by, transport employees and
friends/family (via mobile phone) when it was required during a trip.
Figure 5-1: People - giving and receiving information before/after a trip
Strong ties
both given
and received
Weak ties
Neighbours (transport/travel)
both given
and received
Internet forums
Findings, Phase 1
Figure 5-2: People - giving and receiving travel information during a trip
Strong ties
Weak ties
Family and
friends via
mobile phone
both given
and received
The content of the information which participants reported to have received via word-of-mouth
fell into three broad categories: a general idea which you could then go and check out (“did you
know you can get there by….?”); a specific detail about, for example, routes, timings, costs,
safety, location of bus/train stops, levels of congestion; or a general evaluation of a particular
type of journey, or mode or transport service. Information obtained or offered by word-of-mouth
ranged from the factual (formal) to the very subjective (informal). To illustrate the latter, students
new to the city had been given the following types of advice:
“Not to go on the bus, because that‟s the bus that people are not so nice on, or that
goes through a not very nice area, or don‟t get on that bus because it‟s really filthy. You
get that a lot”. (female)
“it‟s also to do with the area itself, like you get told not to take certain bus routes at
certain times of the night, or if you haven‟t got any way home to get a taxi because it‟s
not safe to take that route because it‟s got a reputation…” (male)
Findings, Phase 1
Although only four of the participants used a bicycle as their primary mode of transport to the
university, roughly half the sample talked about cycling for other trips (including leisure). Wordof-mouth information-sharing, particularly about cycle routes, appeared to be common among
this group; this was explained by the comment that many attributes of a trip by bicycle cannot
easily be gleaned from formal information sources such as maps.
“Little things can make a big difference to how easy a journey is…. Just the traffic, the
wind, how narrow the roads are. The hills.” (male)
“But I think, especially with cycling, it's helpful to know areas to avoid, with lots of traffic
and so forth, to make it a more pleasant journey… It's just nice sharing information. So
it‟s like an informal little network of tips going on perhaps.” (female)
Informal information about safety when cycling on isolated routes was also valued. As well as
obtaining explicit route information, the two female participants who regularly cycled to work
expressed the view that interactions with other cyclists served a motivational purpose when they
were first considering this mode: “it was just nice knowing that it was doable….”
Informal route information was also frequently used to plan car journeys, often in combination
with route-planners, maps and/or Satnav. Several incidents were described where Satnav
directions had conflicted with word-of-mouth information from people who knew the route, which
could lead to issues of which source was considered the most trustworthy:
“It was strange because I was being driven by my partner, and he‟d never been to Milton
Keynes. And I was taking him the way that my parents take me, so it was an
experienced journey. And the Satnav wanted him to go another way, and it created a
certain level of “who am I going to trust more” – the lady next to me, or this thing on the
computer which pretends it knows where we‟re going”. (female)
This participant went on to explain that her preferred route took into account subjective and
personal factors such as “pleasantness”, “nice places to stop” and the location of useful facilities
en route, such as cash machines.
Internet route planners were frequently used, but helpful complementary information could be
obtained though word-of-mouth, such as knowing the “pinch points”, the quieter times to travel,
and pleasant places to stop for a break. For train travel, participants had typically received or
passed on to others information about getting the cheapest tickets, the most convenient stations
for particular destinations and end-legs.
Findings, Phase 1
The reported circumstances in which information was transmitted through word-of-mouth could
be grouped into four categories:
The sharing of experiences and opinions about travel through general interactions,
either before, during or after a trip.
The passive absorption of specific information which crops up in conversation, is not
salient at the time, but may later be recalled if it becomes relevant (and hence
becomes part of active decision-making about that trip).
The active use of specific information which crops up in conversation, and is salient
because the information recipient is planning a similar trip; the information thus
becomes part of active decision-making.
The active seeking or offering of information when an unfamiliar trip is being planned
or information is required during a trip.
i. General social interactions about travel
Firstly, general social interactions about travel were reported to be very common in many social
situations. This was not perceived as actual information-sharing, but rather a matter of everyday
conversation, and might be interpreted as a process of internalising social norms within peer
groups such as the workplace, or among course-mates. Many reported that certain perceptions
about travelling around Bristol tended to arise repeatedly in conversation, notably that the buses
are unreliable and overcrowded, and the traffic congestion abysmal. This sort of conversation
was thought to be a social ice-breaker and a way of contributing to group togetherness (see
5.3.2 for links with social identity theory), rather than something which might actively influence
travel behaviour. However, it also seemed to be helping to form or reinforce participants‟ beliefs
about and attitudes towards particular transport issues. For example, even those who never
used buses in Bristol „knew‟ from what people had told them that bus services were poor and
expensive, and those who never travelled to the university by car knew about the congestion
and parking problems. This kind of discourse was especially evident in the focus group
discussions, during which evaluative statements of this nature tended to attract general support.
For those who did use these modes, these general (negative) interactions seemed to reinforce
what they already knew. Far from encouraging them to think about changing their behaviour,
such conversations appeared to be solidifying it by confirming, for example, that traffic problems
are just something which everyone tolerates. On the other hand, some participants thought that
Findings, Phase 1
awareness of norms of behaviour within their office or classroom environment might cause them
at least to think about their own modal choices, if they realised they were operating outside the
“Maybe I was assuming that there‟s only one way of doing something, then I found out
that for other people, the normal way is doing it x, y or z (…..) and obviously they‟ve got
so much information about that because everybody else does it so they all know. (….) I
might be prepared to think “well maybe”. Reconsider doing it. If something is normal.”
Therefore, it is postulated that general word-of-mouth interactions may be playing a role in
helping to form, reinforce or challenge individuals‟ beliefs and the attitudes arising from them
(particularly their evaluations of particular transport modes). Although the content of such
interactions might not be perceived as „travel information‟ with a direct bearing on cognitive
processes of decision-making, they might be influencing travel behaviour indirectly by affecting
the way in which travel choices are framed, and what types of information are sought, once an
active decision process is initiated in relation to a specific trip.
ii. Passive absorption of non-salient information which “crops up in conversation”
The second category of interactions are those where information about a specific trip was
deemed to have cropped up in conversation, was not salient at the time, but was passively
absorbed and later recalled when it became relevant (i.e. when a similar trip was being made). It
thus became part of active decision-making about a specific trip at a later date. This was
described by several participants as “picking up on other people's experiences”
“I think a lot of the time you talk to people and you hear things from people and you sort
of sense, subconsciously, some of these things register. And then when you're trying to
plan a trip you then think, oh yeah”. (male)
Like the general interactions about travel described above, this process may be conceptualised
as “passive information catching” (see Um and Crompton‟s model, Figure 2-1).
iii. Active consideration of salient information which “crops up in conversation”
In the third, and related, set of circumstances, specific information was once again said to crop
up “in the course of a free-flowing conversation” but in this case was salient at the time because
the information recipient was planning a similar trip. For example, someone might mention a trip
they were planning, or a new job they were starting, and a friend or neighbour would volunteer
information about getting there. Information is relevant at the time and is immediately absorbed
Findings, Phase 1
into an active process of decision-making about a specific trip. This was the most frequently
cited mode of information-sharing:
“Quite often when I mentioned I was going there a lot of people said, oh, I've been there,
it's really nice. Have you thought about going this way or that way, and they‟d volunteer
the information. And so that kind of came through voluntarily from other people once they
knew that I was going there. So that wasn't even something I particularly sought out. It
just came as part of people's reaction to hearing I was going.” (female)
iv. Active seeking or offering of information as part of a cognitive decision process
Finally, information might be actively sought or offered by word-of-mouth, either when
participants were intending to make a specific trip, or when a trip was underway. This type of
„active information search‟, directed towards making a specific travel decision, is the usual focus
of studies of the role of information in behaviour, within frameworks such as information
processing theory and cognitive theories of decision-making (see Section 2.1). Within social
psychology it is consistent with the model of goal-directed behaviour developed by Bagozzi and
colleagues (see Section 3.2). Consistent with the literature on formal travel information-use (e.g.
Farag and Lyons, 2008a and b), informal information tended to be sought to help plan unfamiliar
trips, particularly those which were time-sensitive or involved several interchanges. Word-ofmouth sources were considered to be especially helpful at the local level. For example, in the
context of planning their travel to the University before they first started working or studying
there, two participants said:
“Before I started this job I asked anybody who had any sort of relation either to UWE or
anything in the surrounding area about how they got to work, what route they took etc .
Particularly for driving.” (female)
“When I knew I was coming, I asked lots of people. I suspect, I asked nearly
everybody….. you know if I met somebody who I knew was a cyclist, I think I would have
asked them….” (male)
Pre-trip information was generally sought less frequently when people were travelling with
others, as this was felt to furnish travellers with greater confidence to deal with matters as they
arose; an exception to this was the case of travelling with children, which was felt to require
more pre-planning and could be greatly facilitated by informal information from other parents.
Information was also reported by most participants to have been actively sought (or offered) en
route. For example, information about buses was frequently requested from other passengers
Findings, Phase 1
while waiting at bus stops, especially asking other passengers when buses were expected. This
was particularly the case where real-time information was not available and when buses were
not thought to adhere to „official‟ timetables . One participant remarked:
“For the buses, I pretty much ignore timetables, because they never run on schedule….I
normally ask others at the bus stop.” (male)
As noted above, mobile phones were frequently quoted as a vehicle for actively obtaining wordof-mouth information from friends and family while en route, particularly amongst the younger
participants. This reduced some of the need both for pre-trip planning, and seeking information
from strangers if something unexpected occurred en route.
The perceived role of word-of-mouth information in decision-making processes
Research Question 2.
What is the perceived role of word-of-mouth information in the travel decision process? How do
informal and formal types of information interact?
The analysis of participant accounts in response to this question was strongly interpretative, both
on the part of participants themselves (their own interpretation of the influence of word-of-mouth
on their attitudes and behaviour), and the researcher‟s additional interpretations, regarding, for
example, participants‟ self-perceptions - c.f. the double hermeneutic referred to on p45.
The majority of participants considered informal travel information to be complementary to formal
types of information when used in an active cognitive process of deciding about how, when and
where to travel (levels iv. and iii. above). It was described as “nice to have”, often providing extra
detail which might make a journey easier or more pleasant. Whilst for some it was a useful extra,
others used it as a matter of course, actively „factoring in‟ word-of-mouth advice when making a
new trip. Many said that they used word-of-mouth for reassurance or helping to confirm a
decision about a planned trip.
“So, and then I was speaking to my friend about travelling here, it's just to reassure you,
I think, and help you make up your mind, before making your travel plans. You know, so
they are not essential but they are useful”. (female)
Similarly, many participants gave examples of approaching fellow passengers on public
transport for reassurance (having already checked timetables and variable message signs), or
being asked questions by other passengers. Conversely however, some people spoke of relying
predominantly on word-of-mouth and using formal information only for confirmatory purposes.
Findings, Phase 1
The majority of participants said that if they were forced to rely on either formal or informal
information they would prefer the formal because it was likely to be more objective and easier to
understand. However, others thought they were more likely to follow a personal recommendation
rather than “following the official route”. This was sometimes explained as a matter of
convenience, or of having greater trust in a person with experience compared with an official
information source (both factors will be discussed in Section 5.3.1). For example:
“I generally mistrust formal information when it comes to public transport. Generally you
get a better idea from word-of-mouth”. (male)
Similarly, several participants said they would sometimes prefer to phone a friend for help with
way-finding when they were in an unfamiliar place, rather than relying on formal sources such as
maps. Generally however, it was felt that factual information and informal advice were two
different things which “work very well together”. Examples of these complementarities included
consulting maps and/or journey planners prior to an unfamiliar trip, but also seeking advice from
people with local knowledge about traffic conditions or qualitative aspects of cycle routes.
As mentioned above, many participants believed that information received through word-ofmouth often took the form of a general idea; they would then need to go and check out the
details themselves.
“And if you think, okay, I'm interested in what they're saying, you‟ll then go and check it
out. Because it‟s their opinion, and they've not necessarily got the detail… the informal
(information) would give me an idea of where to look for the formal. So I wouldn't ignore
it, but I wouldn't go on it alone, because I don't think there's enough detail in it”. (female)
However, some participants attributed greater importance to certain pieces of informal pre-trip
information, considering travel advice from family, friends and colleagues to have had a
significant impact on the execution of particular aspects of a trip, such as route, departure time,
station, where to park or stop for a break. Informal information was less likely to have influenced
choice of mode, which was often perceived to be fixed for instrumental reasons (e.g, cost, time,
the only option available, or the need to „trip chain‟). However, word-of-mouth sometimes
affected modal choice when people were travelling abroad, and in some cases these decisions
were influenced by informal information from internet reviews and blogs (perhaps because
people feel less „locked in‟ to particular modal choices when they are outside their everyday
environment). These tended to be consulted only because the participant did not know anyone
personally who had experience of that trip.
Findings, Phase 1
Many participants spoke of gathering information, including through word-of-mouth, considering
different options and making their own decisions. Several spoke of “following my own advice”,
“making my own mind up”, and “knowing what‟s best for me”, in a discourse reminiscent of
individual utility maximisation and rational choice theory. However, new information was thought
to be mixed in with existing knowledge and coloured by personal travel preferences and
established habits, which has a greater congruence with the Theory of Interpersonal Behaviour
(Triandis, 1977 – see Section 2.2). Consistent with information-processing theory, this then
forms part of a feedback loop in which a new travel option is tried on the basis of new
information (which might be formal or informal), assessed on the basis of experience, and fed
back into the process of deciding about future trips.
Interestingly, participants provided more examples of instances where they believed they had
influenced the modal choices of others, than examples of being influenced themselves. This was
particularly the case where people who felt strongly about the benefits of walking or cycling
(especially for fitness) thought they had been instrumental in another person‟s decision to try out
the same mode. Others thought they had generally “put people off” particular transport modes by
recounting their bad experiences, for example of buses, taxis and cycle routes. These instances
provide examples of (perceived) normative social influence, explored in the work of Cialdini and
others (1990). They may also indicate that people assume their influence on others to be greater
than the subjects of such influence might themselves believe.
As discussed in Section 5.1.3 above, processes of general social interaction and the picking up
of non-salient information through conversation (levels i. and ii above) were not considered to
have had a direct effect on final decision outcomes once a trip was being planned, although
participants described a possible influence on their general way of thinking through
subconscious or passive processes of absorbing information. For example, one participant
reported trying out a different transport mode from his usual one following a passing comment
from a course-mate, explaining that the comment had “made him think”, although it did not have
a direct bearing on the final decision, which he described as reaching through an individual and
rational process. Returning to Um and Crompton‟s model (Figure 2-1), “passive information
catching” through general social interactions may be contributing to the formation of beliefs
about different transport possibilities (the “awareness set”) which are later drawn upon when an
active choice process is initiated.
Participants also described general social interactions with people they did not know whilst trips
were underway, usually in the context of problems occurring on public transport. Although this
form of information-sharing sometimes played a role in individuals‟ immediate decisions about
what to do next, it also helped to generate a feeling of „group togetherness‟ (as outlined in
Findings, Phase 1
Section 5.1.3). Complaints about transport were seen as a way of bringing people together and
encouraging them to cooperate in conditions of adversity – described by one participant as the
“Blitz spirit”.
Having discussed participants‟ accounts of the influence of informal information on decision
processes, the next section moves on to the third research question, which explored socialpsychological factors which might help to explain the use and effects of information shared
through word-of-mouth.
The role of social-psychological factors in word-of-mouth information-use
Research Question 3
How do particular social-psychological factors (e.g. social norms, social identity, pro-social
values, trust) appear to influence the use and effects of word-of-mouth information?
Participants were asked whether they believed social considerations such as their relationship to
the other person, commonality of interests, social and personal values relating to travel, or
concern with social norms, affected their propensity to consider, follow up, or act upon word-ofmouth travel information, or indeed their inclination to seek or offer informal information at all.
These areas of enquiry reflected some of the key „social‟ constructs arising from the literature
review of social-psychological theories of behaviour. Of particular interest were those factors
which were thought to render the information provided by one person more reliable and
trustworthy than that of another. The aim was to explore whether social-psychological constructs
such as social and subjective norms, pro-social values, self-concept and social identity emerged
as significant within people‟s own explanations of their thought processes and behaviour (and
those they attributed to others). Although, as with all the Phase 1 data, thematic analysis was
the main method of analysing the transcripts, the thinking in this part of the analysis was
informed by both phenomenological (focusing on meaning) and interpretive approaches – in
particular interpretative phenomenological analysis, outlined in Appendix D.
The majority of participants considered that holding common interests, similarity of
circumstances, membership of a particular group, or closeness of relationship to the person
providing information, would have some degree of influence on how seriously they were likely to
view this information. Central to this was the concept of trust, but there were differing views as to
whether this trust arose simply as a result of the person‟s experience of a particular journey
(„local knowledge‟), or whether it reflected a more deep-seated trust in them as a person, which
Findings, Phase 1
emanated more broadly from their relationship with that person. In many instances, the two were
thought to be intertwined:
“…if I trust the people, I would trust the information…..you become a judge of people, and
some people you trust with information, and other people you wouldn‟t trust with
information” (female).
Similarly, trust was thought to be linked with sharing common interests or attributes:
“I mean, you normally trust people that you feel you have something in common with,
don‟t you?” (female).
In general however, the past experience of the information-giver was thought to make the quality
of their information more trustworthy or reliable, and this was more important than the closeness
of the relationship one had with that person (an „instrumental-reasoned‟ rather than a „social‟
explanation); unsurprisingly, if the information-giver did not have personal experience of the trip
in question, their information was unlikely to be given serious consideration, however close the
relationship or trustworthy they were felt to be as a person. Relating this to the literature review,
it appears that “calculus trust” or “reliability” may be more significant than ”emotional” or
“relational trust” in the domain of travel information (Rousseau et al., 1998; Johnson-George and
Swap, 1982). This appeared to be particularly the case when the content of the advice was more
factual, such as route directions.
“You need to trust the person, and it doesn‟t matter so much if they are close…I‟d just
need to know that they knew, or thought they knew, what they were talking about.”
Several participants mentioned close family members or friends whose travel advice they would
not trust, because they knew them to have different modal preferences, different attitudes to
time-keeping, or simply a poor sense of direction. Several participants remarked that knowing
the person well allows you to evaluate any information or advice according to what you know of
their personality, attitudes and preferences, so that you can apply it to your own circumstances
without necessarily taking it at face value. “I think if you get to know somebody well, then you
can temper what they give you” (male). In considering information provided by a colleague, one
participant commented:
“I would take account of the whole picture of what I knew of her and her opinions, I
suppose, in evaluating any advice or information she gave me. And I think it‟s very difficult
not to really.” (male).
Findings, Phase 1
In this case, the participant believed that the information-giver had different personal norms from
him with regard to environment (c.f. Schwartz, 1977), and he took this into consideration when
evaluating what she told him. If the friend or relative‟s preferences and attitudes did happen to
coincide with those of the information recipient, this would make them seem more trustworthy
than a person they knew less well. In this case, “relational trust” arising from the relationship with
the information-giver serves to reinforce the “calculus trust” arising from that person‟s knowledge
and experience:
“If somebody had the same knowledge and experience and I trusted (that) just as much
as my family member, I would probably still listen to my family member (….) just
because that‟s another good point to add to the final thing altogether”. (male)
Trust in another person‟s knowledge appears to be highly nuanced, even within the specific field
of travel. For example, you may trust a person‟s subjective judgement about ease or comfort, but
not their sense of direction. You may trust in their experience, but not in their ability to remember
and communicate it accurately. Speaking of her neighbour, one participant commented:
“She has no sense of direction, so I wouldn‟t ask her whether to turn left or right
anywhere, but in terms of ease of using various transport, like she uses the train and the
car, and how easy she finds those….I would trust her implicitly.” (female).
Information received through word-of-mouth was also thought to be trustworthy if it were specific
rather than vague, if it were provided by someone local, if it did not appear to be overly directive,
and if the information was provided in a way which implied that the information-giver had similar
travel preferences (even if the information recipient did not know them well). The last point was
particularly important for people reading online reviews, who spoke of looking for „social cues‟ to
judge whether a source was trustworthy.
Similarity and in-group identity
The effect of sharing common interests with the information-giver, perhaps through membership
of a social group, or similarity of circumstances, such as working in the same office, has a
certain resonance with the concept of social identity (Tajfel, 1982, Tajfel and Turner, 1986),
which suggests that people derive a sense of identity from membership of certain social groups
which have emotional significance for them. A number of examples were provided of
information-sharing within a particular group as a way of supporting other members. One of the
participants who cycled regularly commented:
Findings, Phase 1
“I think, because there‟s fewer of us in the work environment…maybe there‟s a kind of
looking out for each other a little bit more than car drivers….”. (female).
One participant remarked that he would be more likely to listen to local travel advice from other
UWE students than from anyone else, because “us students, we‟ve got to stick together”. These
in-group memberships appeared to have some emotional significance for these participants.
However, a sense of strong in-group support versus out-group ambivalence, characteristic of
social identity theory, was not revealed within the interviews or focus groups. This may partly
have been a reflection of the research method and the discourse of individualism running
through the discussions, and because no particular group memberships were salient; these
issues were thought to warrant further investigation in Phase 2. When it came to listening to
specific pieces of advice, one-to-one similarities with the other person (e.g. socio-demographic
similarities, similar attitudes to transport, similar personal norms, or shared interests) appeared
to be more important than belonging to a group.
Social judgement
Social considerations were thought to be less of an issue when people sought or offered on-trip
information from fellow passengers and needed to make a quick judgement about whom to
approach. The main reasons for approaching a particular person in this case were because they
looked knowledgeable or experienced, as well as friendly and approachable.
“You ask the person who looks really confident. And eventually perhaps, you start looking
like that person, rather than the person who keeps looking around and looking at the sign
all the time” (female).
People‟s perceptions of what made them or another person look knowledgeable did, however,
vary. For some, it would be because a person “looked local”, whereas several of the younger
participants said they would prefer to ask somebody older than themselves. According to socialpsychological theory, the process of making quick judgements in this way could be described as
one where participants were drawing on stereotypes: shared beliefs about personality traits and
behaviours of group members (Fiedler and Bless, 2001). This area was thought to warrant
further investigation in Phase 2. Furthermore, some of the younger participants said they were
reluctant to approach strangers for information when travelling in an unfamiliar place, for fear of
appearing vulnerable.
Findings, Phase 1
Pro-social behaviour
Providing pre-trip directions or advice about modal options, for example to visitors to one‟s home
or workplace, tended to be described in instrumental-reasoned terms: “making things run more
smoothly” or “making life easier”. Some people said they would put themselves in the other
person‟s shoes, advising them with regard to what they knew to be their visitor‟s modal
preferences, rather then trying to steer them in the direction of their own. This also applied to
giving directions: i.e. describing a route which was easiest for that person, rather than
necessarily the one they would choose. This could be interpreted as pro-social behaviour,
defined by Bierhoff (2002) as “helpful actions intended to benefit another person, which are not
undertaken through professional obligation”. However, some people said they would tell
someone about their favourite option first, but largely because they knew most about it and
believed it to be „better‟, rather than through attempts to be prescriptive.
Most participants described their motivations for offering information to fellow passengers in
instrumental-reasons terms: e.g. “just being helpful”. When probed, participants spoke of
feelings of empathy: “I know how unpleasant it is to be lost”; and reciprocity: “people have
helped me when I‟ve had a problem” and “Because you know what it feels like, don‟t you, the
other way around”. Both empathy and reciprocity are facets of pro-social behaviour (Bierhoff,
2002). However, most people were reluctant to volunteer advice to others, particularly strangers
en route, unless they were specifically asked, or unless it was obvious that the person was
floundering. Some commented that it is not normal in our society to help people you do not
know. Some people said they would only offer information if they were confident in their own
knowledge. Their degree of helpfulness might also depend on their mood that day or how many
other people were around (you are less likely to help someone if you are in a crowd). The latter
phenomenon is referred to as the „bystander effect‟ in social psychology; experiments have
shown that social inhibition makes people less likely to intervene and help someone in an
emergency situation if other people are present (e.g. Darley and Latané, 1968; Latané and Nida,
Personal norms and values
Relatively few participants said their (modal) travel choices were directly influenced by personal
norms and values (such as pro-social or pro-environmental values), although several indicated
in the course of the interviews that they were indeed concerned about the environmental impact
of transport. Those who did admit to considering environmental issues in their transport choices
tended to say that they were reluctant to impose their views on others:
Findings, Phase 1
“I don‟t actually discuss it because I know a lot of people don‟t want to talk about the
environment. They‟d rather sit in cars on their own. It‟s not up to me to pass any
judgement on that” (female).
Similarly, many stressed that they did not like to feel they were being pushed in a particular
direction by someone “with an agenda”, and that they would tend to find this patronising.
Paradoxically however, people who had used blogs as a source of travel information when
abroad considered this type of information to be less biased than formal sources because:
“You‟re not really being sold it. I‟m not sure if that‟s the case with everyone, but I am
always suspicious on what I‟m being told. And if it isn‟t being sold to me, even if it is being
sold to me, but it is being sold to me in such a way that it doesn‟t feel like it is being sold,
then I‟m more inclined to use it.” (male)
Subjective norms
The perception that „significant others‟ would or would not care about or approve of a particular
travel behaviour (the subjective norm - Ajzen, 1991) was difficult to explore in the context of
word-of-mouth travel information. The aim of the questioning in this area was to explore whether
participants were likely to give information greater weight or consideration if it had come from
someone who they knew would care about their travel choices (or whether there was any
element of implicit approval-seeking in following the advice of significant others). Several
younger participants described examples of acting on information provided by their parents, and
one person said it “felt odd” when she did not take a route suggested by her father. But generally
they did not see parental concern or approval as influential in whether or not they followed this
advice. Sometimes participants did speak of implicit or subconscious seeking of approval when
acting on the advice of a person they knew well. However, this was more likely to be expressed
in instrumental/reasoned terms: “my friend would think I was stupid if I ignored his advice and
then got lost”, than reflecting any deeper sense of seeking the affirmation of the other person.
This is consistent with previous research which has shown the connection between subjective
norm and behavioural intention to be relatively weak (Bagozzi and Lee, 2002, Terry et al., 1999).
However, it is worth noting that some participants implied in open discussion, as opposed to in
response to direct questioning, that the opinions of a spouse or parent for example were indeed
important to them. This raises issues about possible self-presentation factors; most participants
stressed that they were independent-minded individuals, which may have belied any underlying
concerns they had about how others judged them. It is perhaps unsurprising, therefore, that
admissions of approval-seeking rarely appeared within considered responses to direct
Findings, Phase 1
questions. The limitations of direct questioning as a research method are discussed further in
Chapter 6.
Social norms
Some participants expressed the view that enjoying broader social approval was more
significant than the approval of „a significant other‟. This may relate to social norms within a
particular group, such as work colleagues or course-mates. Alternatively, it may be defined as a
process of internalisation, whereby a decision is adopted based on the congruence of one‟s
values with the values of others (Bagozzi and Lee, 2002). However, this tended to be something
to be enjoyed if personal travel behaviour happened to correspond to group norms, rather than
being a factor which might influence an individual‟s behaviour. Everyone was aware of the
„normal‟ way to travel amongst their colleagues and course-mates, but many said that they were
not influenced by these norms personally (in fact, several people spoke of consciously
challenging the norm).
More generally, some participants thought that they were more influenced by information which
communicated injunctive social norms (as opposed to descriptive social norms – see Cialdini et
al., 1990). For example, information which suggested alternatives to flying and also implied that
flying should be avoided for environmental reasons, would be given greater weight than
completely objective information, if this view corresponded with an individual‟s own personal
Some also remarked that they believed people to be influenced by group norms and other social
factors but were unwilling to admit it.
“I think we try not to be influenced by other people and seek approval openly. But I think
underneath we are influenced. It's like, as we were saying, in our office there are quite a
lot of people who walk or cycle. Now, I've never had that - in a previous job everybody
drove, so I never entertained the possibility of walking, or getting a bus, or anything like
that. You just don't think. You get in your car and you drive, because everybody else
does. But I think, because other people do cycle and do walk, then you think, well if they
can do it, then I could, couldn't I? And it‟s this thing of you would get approval and
accepted. It‟s not that it's necessary, it's just again that you have something in
common…….. So I think it does affect you in the decisions you make” (female).
Findings, Phase 1
Figure 5-3 and Figure 5-4 summarise, respectively, the main motivations cited for giving
information to others both pre-trip and en route, and the main social-psychological factors
affecting the use of word-of-mouth information in the two contexts.
Findings, Phase 1
Figure 5-3: Social-psychological factors: motivations for giving information to others
Giving inf ormation to others
inf ormation
En route
inf ormation
Making it easier
f or people
Making it easier
f or people
Sharing local
Not wanting to
impose opinions
Not wanting to
impose opinions
Wanting to
inf luence others‟
Being in the same
Findings, Phase 1
Figure 5-4: Social-psychological factors affecting use of information received from others
Receiving inf ormation f rom others
inf ormation
Experience of
inf ormationgiver
Attitudes to or
relationship with
inf ormation-giver
Who looks
How well
you know
the person
Trust in
inf ormation-giver
Being in the
same situation
as others
Personal saf ety
Similarities with
inf ormation-giver
Findings, Phase 1
Other areas of findings
The previous sections have discussed those findings which relate specifically to the three main
research questions posed in this phase of the study. However, a number of other areas were
covered in the interviews and focus groups which are felt to have an important bearing on the
direct research questions.
Cognitive factors
Some participants said their main motivation for asking for information by word-of-mouth rather
than seeking formal sources was to save time and effort:
“I find it‟s usually easier to ask people who know. I‟m much better at asking people
something instead of looking it up on the computer or finding it in a book or whatever. I
just find it‟s quicker for me.” (female)
In these cases, asking other people for information rather than looking up formal sources
appeared to provide a cognitive short-cut (Tversky and Kahneman, 1974). Conversely, cognitive
factors were sometimes thought to be a barrier to informal information-use, particularly in
relation to word-of-mouth route directions. Consistent with literature on bounded rationality
(Simon, 1959), several participants said they found it difficult to follow verbal directions because
different people perceive and describe things in different ways. Cognitive limitations might also
make it difficult for people to provide accurate word-of-mouth information, sometimes making it
seem less trustworthy as a result:
“I know there's lots of things that aren't on the maps. And I definitely trust that what the
person has in their head kind of better. But I'm not sure I quite trust their ability to
communicate (….) what they are remembering to me”. (male)
“It‟s difficult when you stop to ask people because they know things in a different way
from what perhaps you do. You might be asking for a place and they might notice
something different. You are literally talking about the same thing but you are describing
it in a different (way).” (male)
One participant said she felt inhibited in asking strangers for directions in case she had difficulty
in following them. Some participants were also critical of their own abilities to provide accurate
word-of-mouth directions, to the extent that they sometimes preferred not to give information at
all rather than risk giving the wrong information, even if this made them seem unhelpful.
Findings, Phase 1
However, difficulties in processing traveller information from formal sources were also frequently
referred to, usually in a way which involved criticising particular information sources for being
difficult to understand; common targets of criticism were web-based journey planners for rail and
bus, and simplified route maps for bus and underground which misrepresented distances and
directions, and were therefore deemed to be of little use to people who did not know the area or
the names of stops.
Personality-related factors
The amount of information sought or offered by word-of-mouth also appears to be affected by
personality factors, although there were more similarities than differences between participants
in this respect, which might in part be a reflection of cultural expectations - with regard to timekeeping for example. Virtually every participant said they “hated being late”, and many reported
feeling stressed by time-sensitive journeys (and annoyed with themselves for feeling stressed),
especially in the context of job interviews and starting a new job. This made them more likely to
seek travel information both before and during a trip; the role of word-of-mouth advice in
providing reassurance and reducing uncertainty seemed to be especially important in this
respect. However, several participants commented on the significance of the trip context,
pointing out that they were much more relaxed about trips which were not time-sensitive, and
therefore less likely to seek advice. A further strategy for reducing uncertainty was to do a trial
run of the trip before a job interview or the first day of the new job - virtually every one of the
employed participants (and some of the students) had done this before starting their job (or
course) at the university. This suggests that however much external information has been
obtained, whether it be formal or informal, it does not substitute for personal experience as a
means of reducing uncertainty.
Other personality-related factors explored in the research which might affect the seeking or
offering of information by word-of-mouth were self-presentation and self-efficacy. Differing levels
of concern with self-presentation - the process of controlling how we are perceived by other
people (Leary, 1995) - appeared to have some impact on people‟s willingness to ask others for
advice or to offer it themselves. Whilst most professed themselves always happy to approach
other people for information and advice, be they friends, acquaintances or strangers, for others it
was something they forced themselves to do because they would feel more foolish if they failed
to ask and then made a mistake.
“I think it is natural to be concerned to ask and show you don‟t know, but I think I‟ve
worked on that and I‟m OK about it and I think I‟d ask (…) if I could.” (male)
Findings, Phase 1
Several participants expressed concern about not wishing to appear a „know-all‟ by offering
unsolicited advice to others, and many said they would be hesitant about approaching strangers
to give them information unless it was very clear that they were in difficulty, although it was
outside the scope of the current research to probe into whether, once again, this was a matter of
individual personality or cultural norms.
Participants were also asked to consider their word-of-mouth information-seeking behaviour in
relation to self-efficacy: belief in one‟s own ability to reach a specific goal (Stroebe and Jonas,
2001). Unsurprisingly, everyone considered themselves capable of planning or executing a
journey without input from other people, although in many of the specific trip examples provided,
participants believed that it would have been more difficult, time-consuming, or even impossible
to do this without the benefit of information provided by others. Self-efficacy is similar to
perceived behavioural control (Ajzen, 1991) and in relation to travel information-use might be
regarded as context-dependent rather than a stable feature of personality.
Past (personal) experience
The importance of past experience in influencing attitudes and behaviour is well documented,
featuring in both the information-processing theories and the more complex of the (socialpsychological) behavioural theories reviewed for this thesis. In the former it is the locus of the
„internal information search‟ which precludes or obviates the need for an external information
search, then feeds back into future information searches once a behaviour has been enacted. In
some of the latter, frequency and/or recentness of past behaviour is a moderator of behavioural
intention (e.g. as a precursor to habit in Triandis‟ theory of Interpersonal Behaviour) or of trying
to act (e.g. Bagozzi‟s Theory of Trying). In the context of this study, past experience was found
to affect informal information-use in a number of ways. Firstly, some participants appeared less
likely to seek informal information or word-of-mouth reassurance from others as a general rule,
even during or prior to an unfamiliar trip, because they regarded themselves as experienced
travellers, which gave them a high degree of self-confidence – or self-efficacy - with regard to
travel (this group also said that they were frequently asked for information and advice by others).
Secondly, consistent with information-processing theory, past personal experience of a similar
trip reduced the perceived need for external information of any type. Personal experience clearly
took precedence over what participants had learned from others about their experiences. One
participant who had cycled to work in her previous job remarked:
“I‟ve lived in Bristol for quite a while, so when I got the job I knew where UWE was and I
knew what Bristol public transport was like. So I was pretty sure that I was going to drive
Findings, Phase 1
in (….). and then (…) one of my colleagues said actually there‟s a really good bus
service down to where I live. And I‟ve never even looked.” (female)
However, in most cases where unfamiliar or time-sensitive trips were being actively planned, or
were underway, different types of information tended to be used alongside personal experience
of similar trips in the evaluation of different options. Finally, as mentioned above in the context of
reducing uncertainty, most participants reported using personal experience to aid trip planning in
the sense of doing a trial run of new, time-sensitive trips. Some also spoke of simply trying out a
new travel behaviour rather than employing an extensive information search.
As well as past travel experience, there was also some discussion of past experience of
information-use. For example, many participants said they were “loyal” (perhaps out of habit) to
a tried and tested information source, such as a particular route-planning website, and some
said they would only switch to another source if it were recommended by someone they knew.
Similarly, some participants said that because they had generally had a good experience of
receiving word-of-mouth travel advice (in the sense that it had been reliable and useful), they
would now always include word-of-mouth sources when seeking travel information.
Instrumental factors
In the travel behaviour literature, instrumental factors usually refer to the practical costs and
benefits of a particular trip, such as monetary cost, journey time, ease, flexibility, reliability,
protection from the weather, or health and fitness. These are distinguished from psychological
factors such as habit, and symbolic-affective factors such as self-expression, status and
autonomy, and are often associated in the literature with the „reasoned‟ motives for modal choice
(e.g. Anable and Gatersleben, 2005; Kenyon and Lyons, 2003; Steg et al., 2001). Previous
research, particularly in transport psychology, has highlighted the tendency for people to explain
their travel behaviour first and foremost in instrumental-reasoned terms (see Steg et al., 2001).
It is perhaps no surprise, therefore, that the participants in the current research made frequent
reference to instrumental factors associated with their general travel behaviour (particularly their
modal choices) when considering the specific role played by informal information and word-ofmouth:
“I just take the car because it‟s easier. It doesn‟t relate to any information about what
people do.” (male)
Findings, Phase 1
Referring to general interactions with her course-mates about travel to the university, one
participant remarked:
“I think those are part of the conversations, but because I‟m limited in the modes I‟ve got
anyway, they‟re not going to sway the decision. (…) So I think certainly anecdotal
experiences are part of that information. But how much I listen depends on how much
choice I‟ve got either side of their advice anyway. (female)
Clearly, instrumental factors are understood to exert a major influence on whether travel
information is sought or listened to in the first place (for example, only seeking bus times if you
know there is a direct service), as well as on the final travel choice. As one participant said when
explaining why she had disregarded a suggestion to commute by bus:
“I might want to go somewhere else on the way home (…). So if I was forced to catch
the bus, it wouldn‟t fit in with my lifestyle as well. So even though I knew it would
probably be fine and I‟d had this positive recommendation, I just don‟t actually think it
would fit.” (female)
Interest in web-based informal travel information
In order to provide a bridge between this stage of the research and the applications phase which
was to follow, the interviews and focus groups briefly touched on the participants‟ experience of
and enthusiasm for “electronic word-of-mouth”, such as web-based travel reviews. As reported
previously, some were enthusiastic users of websites such as www.tripadviser.com for holiday
travel, and were experienced in looking for social cues when deciding whether or not to follow up
the tips and advice provided by contributors – i.e. whether the contributor might have similar
likes and dislikes to them. Others had consulted independent travellers‟ blogs, found on Google,
for specific items of travel information when travelling abroad, and reported finding them useful
and more trustworthy than official transport-operator web-sites because they felt that they were
not being „sold‟ a particular travel option. However, this mechanism generally appeared to be
second best to actually knowing someone who had been to the same place.
As mentioned previously, some participants used internet forums connected with a hobby (e.g.
long distance running) to obtain UK travel information relating to specific events. One person
had found a blog which provided information about getting cheaper rail fares, and a few people
had appreciated the informal information added by cyclists to www.bristolstreets.co.uk (but
thought there was room for improvement). Generally, however, participants had not considered
using the internet as a source of informal information for local/national travel, mainly because
Findings, Phase 1
they were not aware of any such sources, although many said they thought this was a good
idea, and would consider using it if they knew where to look (whilst a small minority were
adamant that they would never do so). Most said they thought that a web-based repository of
local knowledge/experience of travelling to UWE would be useful, although there would probably
be issues relating to ease of use, and whom to trust. Comments on such a system included the
it would need a critical mass of users to be effective (otherwise it might be neither
comprehensive nor up-to-date enough to be genuinely helpful);
relevant information would have to be easy and quick to access; for example it could be
ordered by the different areas people would be travelling from;
it would need to be recommended by someone you knew in order for you to trust it;
it would need to be contained within a particular community (e.g. UWE students),
because you would be likely to trust fellow students more than „strangers‟;
knowing that contributors had experience of a similar journey to you would give it
it might give people more options and ideas about travel options;
cycle routes and advice about cycle facilities would be especially useful, as would hints
about reliability of buses.
These suggestions were used to help select a case-study for Phase 2 of the research.
Chapter Summary
The exploratory research revealed the value attributed to informal advice obtained from social
contacts with first-hand experience of a particular trip, and its role in improving awareness of
different travel alternatives and/or improving the trip experience. The circumstances in which
informal information was acquired or transmitted were found to take three main forms: general
social interactions about travel; passive absorption of information about specific trips; and the
active seeking or offering of information during the planning or execution of a trip. General social
interactions about travel (for example, appraising the experience of using a particular transport
mode), whilst not necessarily perceived as travel information, appeared to be influencing beliefs
and attitudes, and shaping the psychological context in which travel choices might later be
made. More specific conversations about particular trips often involved a process of passive
Findings, Phase 1
information absorption, through which items of information were stored and later retrieved if a
similar trip was being planned. Finally, at the stage of planning or undertaking a trip, information
obtained through word-of-mouth was thought to have played a complementary role to formal
information in the process of active decision-making, and was believed to have directly
influenced trip details (such as route or departure time), but was less likely to have affected
participants‟ modal choice.
„Local knowledge‟ was deemed trustworthy primarily because it was based on the other person‟s
direct experience, but trustworthiness could also be improved by social and psychological
factors such as the degree of familiarity or similarity with the information-giver, or, to a lesser
degree, identifying with other members of a particular group. In terms of providing informal
information to others, factors such as empathy and reciprocity, both facets of pro-social
behaviour, frequently appeared in participants‟ accounts. These themes were explored in
relation to all the common forms of everyday transport, but informal cycling information,
particularly about cycle routes, began to emerge as an area of particular interest as the research
evolved. Preliminary enquiry into the potential for informal information-sharing using the internet
(harnessing “digital word-of-mouth”) to enhance the planning of utility trips revealed an interest in
a local, community-based information system in which other users‟ information could be relied
upon both because of their first-hand knowledge and through a sense of belonging to the same
Discussion and transition to the next phase
Chapter 6 : Discussion and transition to the next phase
By using interviews and focus groups as a research method, the exploratory study encouraged
participants to reflect in detail on their use of informal and word-of-mouth information, and its
perceived role in travel decision processes, setting this in the context of their own real-life
experiences. As well as elucidating the perceived influence of word-of-mouth on actual travel
behaviour, it also shed light on its influence on some of the processes which might precede a
behaviour, such as the internalization of norms, and the formation of beliefs, attitudes and
intentions. Unlike structured methods of obtaining data on travel preferences, choices and
antecedent constructs, this methodology cannot demonstrate statistically significant relationships
between behavioural constructs or prove or disprove theories. What it can do is explore people‟s
understanding of some of the behavioural assumptions expressed in theories, and help us to
interpret meanings which may emerge through open discussion (based on an interpretive and
phenomenological epistemology). From this perspective, Bandura‟s social learning theory (1977)
provided a useful framework within which to conceptualise the research participants‟
explanations of the role of word-of-mouth information in everyday travel. They described a
process of “picking up on people‟s experiences” and benefiting from other people‟s “local
knowledge” through word-of-mouth, and in turn, sharing their own travel experiences to facilitate
other people‟s journeys; in social learning terms, it allowed them to learn vicariously from other
people‟s experiences. Word-of-mouth communication thus provides a medium through which
social learning can occur.
The role of word-of-mouth in transmitting social or group norms also stood out in this study. The
findings implied that norms might be transmitted more effectively through a process of passive
internalization rather than of active compliance with „significant others‟ (subjective norm) or
identifying with the „in-group‟ (social identity) (Bagozzi and Lee, 2002). In the domain of travel
behaviour, people may regard norms as exerting a stronger “informational influence” than a
“normative influence”. The former is described as “influence based on accepting the information
obtained from others as evidence about reality”, and the latter as “influence based on
conforming to the positive expectations of others” (Deutsch and Gerard, 1955). Although the
original focus of this research was the influence of informal information on decision-making, the
initial analysis of the findings suggested that word-of-mouth may play just as important a role (if
not more so) prior to the initiation of an active decision-making process, when beliefs and
attitudes are being formed through “passive information catching” (Um and Crompton, 1990).
When active information-search begins, key choices such as mode have often already been
made, or are highly susceptible to other factors, particularly instrumental ones. Whereas general
Discussion and transition to the next phase
interactions and topics which happen to crop up in conversation may be difficult to ignore, even if
they are given little thought at the time, information about specific trip details may never even be
sought, and therefore can be ignored, if the traveller already has a negative attitude: why find out
what time the bus leaves if you already „know‟ from your colleagues that the bus service is
terrible? It might be better to check out another mode which has received a more positive
With regard to the cognitive process of decision-making about travel, a discourse reminiscent of
the individual utility maximiser emerged from participants‟ explanations of their own decision
processes. This was interesting in the light of the criticisms of this model from the fields of
psychology and sociology which emerged in the literature review. Consistent with literature on
motives for modal choice (e.g. Steg et al. 2001, Kenyon and Lyons, 2003) participants tended to
describe their decision processes in instrumental-reasoned terms; different sources and types of
information were evaluated and a „rational‟ decision reached based on what suited them best as
individuals. It is perhaps unsurprising that people wish to see and project themselves as rational
decision-makers bearing in mind how deeply the concepts of individual self-determination and
rational thought are embedded within Western culture, and nowhere is this more likely to be
evident than within a university, from which the research sample was drawn. A different type of
discourse might have emerged had these questions been posed to people belonging to a less
individualist and more collectivist culture. Triandis (2001) offers a social-psychological
perspective on individualism and collectivism:
“In individualist societies people are autonomous and independent from their in-groups;
they give priority to their personal goals over the goals of their in-groups, they behave
primarily on the basis of their attitudes rather than the norms of their in-groups, and
exchange theory adequately predicts their social behavior.” (p.909)
This he contrasts with collectivist cultures where:
“people are interdependent within their in-groups (family, tribe, nation, etc.), give priority
to the goals of their in-groups, shape their behavior primarily on the basis of in-group
norms, and behave in a communal way.” (p.909)
Notwithstanding the preponderance of rational and individualist accounts of behaviour, the
findings have highlighted a variety of other factors identified in behavioural theory, such as
cognitive shortcuts, past behaviour, and the attitudes and experiences of, as well as concern for,
other people. Because of the focus on social-psychological mechanisms within this study, the
role of other peoples‟ attitudes and experiences, and the propensity to help others, are of
Discussion and transition to the next phase
particular interest. A key topic to emerge was the level of trust which could be placed in other
people and the information they provided: what kind of factors rendered one person more
reliable as a source of information compared with another? The concept of “calculus trust”
appears to be particularly important in the travel decision process. Thus, the most important
factor rendering information trustworthy was deemed to be the direct personal experience of the
information-giver. However, "relational" or "emotional" trust also appears to be significant,
particularly in the realm of pre-trip planning, when knowing the information-giver better makes it
easier to evaluate the information they are giving you. This is where social factors such as
familiarity, approval and similarities between information-giver and recipient might have a role in
increasing the perceived reliability of the information. With regard to offering informal travel
information to others, pro-social behaviour, typified by feelings of empathy and reciprocity,
appeared to underlie some of the motivations for doing so, particularly when it came to helping
strangers during a journey.
Returning briefly to the research method, it was sometimes difficult to identify the mechanisms of
social influence which occur through word-of-mouth; some participants said this process
occurred at a largely sub-conscious level. Again, this is consistent with research showing that
people tend to rationalise their behaviour, and this may be compounded in an interview situation,
where people may be concerned with self-presentation (Leary, 1995). It should also be
acknowledged that many of the interview questions were framed in a way which probed the role
of informal information in active/cognitive decision-making, reflecting the original
conceptualization of the research, and this might have introduced an element of interviewer bias
which inadvertently guided responses towards „rational‟ accounts. Of course, careful analysis of
what people say can lead to interpretations of their attitudes and motivations which do not
necessarily correspond with their own reasoned explanations. What is said spontaneously or „off
the cuff‟ may differ from what is expressed within a considered response to a direct question.
This led to the conclusion that different research methods, perhaps involving direct observation
of word-of-mouth behaviour, might therefore prove helpful in developing the research in the
second empirical research phase.
Whether the participants in this study were genuinely rational and individualist in their thinking
with regard to travel (albeit influenced by others to varying degrees), or simply wished to project
themselves in that way, it is useful to consider the implications for transport policy and practice,
and travel information in particular. Behavioural change measures such as travel plans may
benefit from a greater understanding of word-of-mouth information transfer, and how this might
relate to the spread of new travel behaviours within communities through a process of social
influence, although the present research has highlighted that many people may be resistant to
Discussion and transition to the next phase
suggestions that their travel choices are influenced by others. However, no particular group
membership was salient during the Phase 1 interviews (word-of-mouth information-sharing was
discussed in a variety of contexts), whereas group identification might be expected to be
stronger in contexts where social similarities are emphasised – such as within an „online
community‟, or among work colleagues commuting via the same mode of transport. An issue
which was thought to warrant further investigation in Phase 2 was, therefore, whether social
influence might be more strongly observed in a small group context where membership of this
group might be salient.
Implications for travel information provision
With regard to the provision of travel information, it appears that informal information and wordof-mouth processes may influence travel behaviour in complex ways, from the passive
absorption of other people‟s experiences and behavioural norms (influencing beliefs and
attitudes) to the active consideration and use of specific items of information (influencing aspects
of travel behaviour). In the process of active information search (or active offering of information)
it is equally clear that informal information and word-of-mouth delivery mechanisms complement
but cannot replace formal types or sources of information. Nor is it possible to isolate information
from other factors as a motivator of particular travel behaviours. The findings showed that
informal information was considered alongside formal information and weighed up with factors
such as past experience and instrumental opportunities and constraints when part of a rational
decision process. Social-psychological factors were found to affect important aspects of word-ofmouth information behaviour: whom to ask for information and why; why to give information to
others; what makes it trust-worthy; and whether to act upon it, although the final travel decision
was nearly always deemed to be an individual one. No population segment stood out as having
a particular propensity to use word-of-mouth – it depended on personality and experience rather
than factors such as age, gender, family situation, employee or student status. In terms of
transport modes, there were strong indications from the half of the sample who regularly cycled
as one of their transport modes, that word-of-mouth was a particularly useful source of cycling
Figure 6-1 brings together some of the main features of the (pre-trip) travel decision process,
highlighting the role of informal information (encompassing both informal content and delivery
through word-of-mouth) within the overall process, and the main social-psychological factors
which were found to affect its use. It might be over-ambitious to expect advanced traveller
information systems to address all parts of this process. As information systems tend to be
consulted when a trip is being actively planned, the second (applied) phase of the research was
Discussion and transition to the next phase
intended to focus on features of the lower part of the model, particularly the social-psychological
factors (lower left) affecting informal information-use within a web-based case-study system.
However, it was considered important that the second research phase should also remain open
to the possibility that passive information absorption and background normative influence
(depicted in the top section of Figure 6-1) might also occur through the use of an information
system which allows sustained social interaction amongst users.
The following section draws out the main findings relevant to advanced traveller information
provision and suggests the areas for further examination in Phase 2 through the formulation of
preliminary questions. These guided the development of the detailed Phase 2 research
questions, which are set out at the beginning of Chapter 7.
Discussion and transition to the next phase
Figure 6-1: The role of word-of-mouth information in the pre-trip travel decision process
General interactions
about travel
Social norms
of travel
absorption of
inf ormation
“Crops up in
Seek information
Initiation of trip planning
Active seeking
of inf ormation
of info-giver
Attitudes to
How well you
know them
Discussion and transition to the next phase
Moving to the next stage: identification of areas for study in Phase 2
This section outlines some of the main findings from the exploratory research phase, and the
further questions which emerged from them with regard to applications in the field of advanced
traveller information systems. This guided the development of detailed research questions for
Phase 2, which will be set out in the next chapter, and the selection of an information system
case-study in which they might be answered.
Box 6.1
 Informal types of information from other people are thought to provide a useful supplement to
formal travel information by improving awareness of and knowledge about different travel
alternatives and/or improving the experience of a journey. „Local knowledge‟ based on
personal experience is especially valued.
 Informal travel information is conveyed extensively through face-to-face word-of-mouth
among people who know one-another, although electronic word-of-mouth (interactive
websites, web-based discussion forums etc) is sometimes used as an alternative, especially
for travel abroad. Participants expressed interest in the idea of a local, web-based source of
informal travel information, where staff and students at UWE could share tips and
Question: How would people respond to and use a local, web-based source of informal
travel information?
Box 6.2
 For informal information to be considered useful, it must be deemed trustworthy, and to
be trustworthy it must first and foremost be based on the direct experience of the
person providing it. However, perceived trustworthiness is sometimes also affected by
social-psychological factors such as:
familiarity with the person providing the information: knowledge of their personality
and preferences allows you to interpret better what they say and adapt it to your
own circumstances.
Discussion and transition to the next phase
Box 6.2, continued
similarities with the other person: information may be given greater credence if
provided by „someone like me‟, whether that be in terms of socio-demographics,
attitudes to transport, personal norms, levels of fitness, or shared interests.
norms of travel behaviour: information may be given greater consideration if it
reflects norms of behaviour within a peer-group.
Different types of trust may be required for different types of information. For example:
for geographical information such as route directions, trust in the person‟s cognitive
abilities and sense of direction.
for more subjective information such as an assessment of the ease or pleasantness
of a route or mode, trust that a person would „see things in the same way‟ (here the
above-mentioned social-psychological factors play a stronger role).
When informal information is read on the internet, familiarity with the person posting the
information is likely to be lacking - unless this occurs within a virtual community of
regular contributors, or a physical community such as a neighbourhood or organisation.
Instead, people may look for similarities with the information provider („social cues‟) and
use this to assess the reliability of the information. They may also absorb the norms of
behaviour in a particular travel context.
Question: How might these social-psychological mechanisms (particularly trust) function in
the context of a web-based system for informal information-sharing?
Discussion and transition to the next phase
Box 6.3
Pro-social behaviour may motivate people to provide information to others, putting
themselves in the position of the recipient. Giving information to people you already
know may be a matter of wanting to share your experiences, or wanting people to make
similar choices to your own because you consider them to be somehow better (although
in general people do not want to appear directive).
Question: Are people more likely to be helpful to people within their own community? Would
they share more information about local travel, if they could, for example through an online
system? Why might they not behave pro-socially?
Box 6.4
 Although in-group positivity versus out-group negativity (social identity) did not emerge
strongly from the findings, there were indications that some people who cycle might
experience a sense of in-group identity.
Question: Would group identification emerge more strongly if people were sharing
information within a group of „similar others‟ (a reference group), and would a group of
cyclists provide a suitable context in which to study this?
Although cycling was the primary commuting mode for only four members of the original
research sample, it transpired in the interviews that half the participants cited cycling among
their means of local transport. Word-of-mouth information-sharing, particularly about cycle
routes, was reported to be particularly important for cycle journeys, compared with other
transport modes; this was explained by the view that many features of a cycle trip cannot easily
be obtained from „traditional‟ information sources such as static maps or even online cycle
journey planners . Some evidence was also found in the literature for an expressed need
amongst cyclists for better route information; for example, in the US, Priedhorsky et al.(2007)
identified an „unmet need‟ amongst cyclists for a comprehensive and up-to-date web-based
Discussion and transition to the next phase
information resource, and “personalized ratings of byway bikeability” (i.e. personalised
information on the comfort of cycling along a particular road or path).
The present research also found that whilst many of the features reportedly discussed through
word-of-mouth might be described as instrumental, concerning matters such as topography,
traffic volumes and infrastructure, it appeared that information shared through word-of-mouth
about cycle trips also touched on more aesthetic and sensory aspects of the actual experience
of cycling. Informal information about personal security when cycling on isolated routes was also
valued, whilst interactions with other cyclists had also served a motivational purpose for some
people when first considering taking up cycling. Consistent with Skinner and Rosen‟s study of
social identity amongst cyclists (2007), some participants experienced a degree of in-group
identity with other commuter cyclists within their organisation, expressed through a concern with
“looking out for one another”. Thus, it appeared that route-sharing amongst current and potential
cyclists was not limited purely to the transfer of instrumental information (cycling as a means of
transport to get efficiently from A to B), but could also involve the communication of social and
psychological meanings between participants in the interaction. This raised the question of
whether they might, in turn, be influencing one another‟s attitudes and behaviour through the
medium of such interactions (for example, gaining confidence about cycling due to a sense of
mutual support). Therefore, information-sharing among existing commuter cyclists, and those
considering it, appeared to offer a fruitful context for the study of word-of-mouth processes in
greater depth.
The subsequent Phase 2 case-study was also devised in the context of rapid developments in
“digital word-of-mouth” (Dellarocas, 2003) and in particular the diffusion of user-generated
content via the internet. This has vastly expanded the realm of word-of-mouth information
diffusion from one-to-one communication (often face-to-face) to electronic “one-to-many” (Godes
et al., 2005). When asked for their opinions on this topic during the exploratory research, nearly
all participants had expressed an interest in the idea of a web-based source of local, informal
travel information, where users could share hints and travel advice with one another. Thus, the
idea of developing an online environment in which cyclists in a particular location might share
their knowledge of routes and other cycling–related issues, communicating with one another by
means of an interactive map, began to take shape.
Building on the Phase 2 findings, it was anticipated that an exploration of a case-study
information system might generate an in-depth understanding of the use and effects of informal
travel information within an applied context, with specific reference to the role and relative
importance of the different social and psychological factors under study: for example, the types
of information most frequently consulted, whether it was thought to influence beliefs, intentions
Discussion and transition to the next phase
or behaviours, and how far factors such as trust affect the way it is used. Building on Figure 6-1,
a key aim was to develop a more specific model showing the role of word-of-mouth and
explanatory social-psychological mechanisms in the context of travel information-use. In addition
to theory development, the final stages of the PhD study were also intended to generate
practical outputs in the form of reflections about the incorporation of specific „social design
features‟ - if appropriate - into other types of advanced traveller information systems, such as
web-based journey planners currently offering only formal information.
Chapter Summary
This chapter has set out to form a bridge between the first and second empirical phases by
discussing some of the main findings so far and their relevance to the field of advanced traveller
information. A model was developed to show where, and how, word-of-mouth information may
play a part within the complex process of pre-trip decision-making, and how this role may be
shaped by social as well as individual factors. Consideration was also given to methodological
issues, particularly the advantages and limitations of the interview method, and it was suggested
that observation of actual behaviour might complement data obtained through participants‟ self
reports. Further areas of enquiry emerging from the Phase 1 research were also identified, and
reasons presented for choosing word-of-mouth cycling information as a focus of study for Phase
2. The next chapter will present the detailed research questions and methodology for the second
phase of empirical research.
Empirical Research, Phase 2.
Cycology: a traveller information
Methodology, Phase 2
Chapter 7 : Methodology, Phase 2
This chapter begins by defining the detailed research questions for Phase 2, then provides
contextual information relevant to the selected case-study by outlining some contemporary
developments in the field of informal information-sharing and online route-planning for cyclists.
Consideration is then given to methodological issues concerning the case-study as a research
approach, and the specific methods of data generation are set out. The selected case-study
information system is then introduced, and the operation of the study described. Finally, two data
analysis techniques are described which were used in addition to the methods employed for the
analysis of Phase 1 data.
Phase 2 Research Questions
In this section the remaining two main research questions are reiterated, and a number of
detailed sub-questions defined. The findings from Phase 1 about key facets of informal
information-use in a general context, the perceived role of such information in the complex
picture of travel decision-making, and underlying social-psychological factors, were now
transferred to a case-study scenario in order to draw the exploratory findings into the realm of a
specific transport application. A qualitative approach continued to be employed, allowing an
openness to the possibility that the general will not necessarily translate directly to the specific,
and that new themes may emerge which were not apparent in the exploratory phase. Although
the research had now moved from a general to a specific context, there remained a need to take
a grounded approach which observes what happens rather than measures pre-determined
concepts. Nonetheless, a number of detailed research questions were defined, and are provided
Research Question 4.
How do word-of-mouth processes and informal information-sharing operate within an
emerging web-based system incorporating user-generated travel information?
4.1 For what types of informal travel (cycling) information is the case-study system used?
Categories of information emerging in Phase 1 were: options and ideas; specific factual
detail; and subjective advice. Consideration would also be given to ways in which informal
information was combined with formal information in the case-study system.
4.2 What word-of-mouth processes (e.g. patterns of interaction) can be observed within this
small group context?
Methodology, Phase 2
4.3 What influence (if any) has word-of- mouth information exerted on participants‟ attitudes,
intentions and travel behaviour within the case-study?
Phase 1 findings had shown that information obtained through word-of-mouth was reported
to have influenced participants‟ trip details but not more „strategic‟ decisions such as modal
choice. Normative information obtained through general conversations was thought to have
had no direct effect on trip choices, but may have influenced general attitudes. How far
would this be apparent within the case-study system?
4.4 How do social-psychological mechanisms appear to function in this online context?
A number of more specific areas of interest arising from the Phase 1 exploratory research
were identified for further exploration:
4.4.1 Obtaining word-of-mouth information from others:
In Phase 1, trust was found to be a key factor affecting informal information-use. Whose
information did people trust in this online environment, and why?
Relational trust:
 Familiarity (if you know someone off-line as well, or if you come to know them
through regular online interactions)
 Similarity (e.g. sharing socio-demographic characteristics, similar attitudes to
transport, personal norms, levels of fitness, or shared interests)
OR social proximity through being within the same community (e.g. within an
organisation or locality)
Calculus trust:
 How would this interact with an „instrumental-reasoned‟ trust in the other
person‟s experience?
 What might make information seem less trustworthy?
Methodology, Phase 2
4.4.2 Giving information to others:
Why did people contribute information to the case-study system? The following issues from
Phase 1 would be explored:
Pro-social behaviour/ reciprocal altruism/social exchange
Are people motivated by their membership of an in-group (e.g. within an organisation or
What might prevent people from posting information?
Research Question 5.
How might information-sharing amongst travellers be applicable to advanced traveller
information systems (ATIS) more generally?
5.1 Could advanced traveller information systems which currently offer only „formal‟ types of
information benefit from the integration of „social design features‟ (types of usergenerated content incorporated into Web 2.0 systems), and if so what specific form
could these take?
5.2 To which modes of transport (beyond cycling) might information-sharing be most
5.3 What features of a „user community‟ might be conducive to word-of-mouth informationsharing (e.g. group size, composition, types of user).
Following a brief section placing the second phase of the research within the context of current
developments in the field of Web 2.0 cycling information, the remainder of this chapter sets out
the methodology employed to answer these questions.
Context: Digital word-of-mouth, online route-planning and cycling
The area of interactive cycling information was introduced briefly in Section 1.3.3, but this
section provides a little more detail in order to explain the context in which the case-study was
selected and developed.
Web-based discussion of cycling issues in the UK can already be found in abundance on
specialist online forums hosted (to name but a few) by local cycling campaigns and the national
cycling advocacy organisations, personal blogs, and social networking sites such as
http://morvelo.cc/. Online cycle route planners are also increasing in number and accuracy.
Pioneering work in this area was carried out by the Cambridge Cycling Campaign at
Methodology, Phase 2
www.camcycle.org.uk, which released a route planning system in 2006, illustrated with a wealth
of photographs uploaded and tagged by users (over 25,000 in number at the time of writing).
This map uses geographical data from www.OpenStreetMap.org, the free digital map launched
in 2004 and created by users worldwide. In response to the considerable interest generated by
the Cambridge journey planner and photomap, its originators have now developed a national
system: www.cyclestreets.net, with increasingly comprehensive coverage across the UK. A
further example of a sophisticated, interactive map created by the „user community‟ is provided
by Camden Cycling Campaign at www.camdencyclists.org.uk; here contributors have plotted
both utility and leisure routes around Camden on a Google map, whilst route planning is
available through a link to www.cyclestreets.net. Parallel, „top-down‟ developments in online
cycle journey planning have occurred regionally, such as the Transport for London cycle route
planner, and at the national level through a partnership between Transport Direct, Cycling
England and Ordnance Survey. At the time of writing, Transport Direct‟s new cycle journey
planner at www.transportdirect.info covers 32 English cities/areas.
In the field of utility cycling there has been little convergence so far between the sharing of
„social information‟ (personal accounts of the actual experience of cycling a route), such as one
might find within an online discussion forum, and geographical information in the form of routes
drawn on maps or described by an automated journey-planner. Many cycling websites feature
both interactive maps and discussion forums, but on separate pages of the site. Moreover,
systems such as the www.cyclestreets.net photomap, on which users may add comments to the
photographs they upload, do introduce a more personal element to the routes which they
illustrate. However, the maps themselves do not offer obvious opportunities for social
interactions between users about the routes they display. Therefore, in order to be able to study
social and psychological aspects of online route-sharing amongst cyclists, it appeared that an
innovative approach would be needed, and a bespoke case-study system developed which
would enable (and encourage) social interactions amongst the participants. This system was
developed by Toby Lewis as part of www.bristolstreets.co.uk, and given the name Cycology.
Subsequently, a route-planning website incorporating cyclists‟ comments has come online in
Minneapolis-St Paul, USA (www.cyclopath.org), and this will provide a source of comparison for
the recommendations arising from Cycology in Chapter 11.
The case-study system will be described in Sections 7.6 and 7.7 of this chapter, but first,
consideration is given to methodological issues surrounding the use of case studies. The
specific research methods which were employed within the case-study framework are then
Methodology, Phase 2
Methodological framework: an „experimental case-study‟
This thesis takes the position that the case-study is a research strategy (e.g. Robson, 1993),
within which different research methods may be employed, rather than being a research method
in itself (a position argued by Yin, 2009). Robson (1993) defines case-study as:
“ a strategy for doing research which involves an empirical investigation of a particular
contemporary phenomenon within its real life context using multiple sources of
evidence” (p51).
Blaikie (2000) describes a case-study as “an umbrella term for a family of research methods
having in common the decision to focus on inquiry around an instance” (p215). Yin (2009) adds
that case studies are appropriate when understanding of the phenomenon encompasses an
understanding of important contextual conditions, because they are highly pertinent to the
phenomenon of study. This strategy was relevant to the second phase of empirical research
because a detailed understanding of the use and effects of a „real-world‟ interactive information
system was sought, but to do so required an understanding of the contextual conditions of its
use. Understanding the context (such as the environment in which it was used and the travel
opportunities and constraints facing its users) was especially important for the interpretation of
the findings regarding the behavioural effects of information-sharing, and underlying socialpsychological factors. The use of several different data collection methods was thought
important as a means of gaining this understanding – another facet of the case-study approach
Where the second empirical phase differed from conventional conceptions of the case-study was
that the chosen information system was not „naturally occurring‟ – it was not an existing website
which people were already using in a natural setting. Had such as online system existed where
social interactions were of sufficient quality to answer the research questions, and if online
contributors could be interviewed or surveyed, this would have qualified clearly as a case-study.
However, as noted in the previous section, no such system had been identified, which meant
that a system needed to be specially designed and research participants recruited, creating, in
essence, an „artificial‟ research-manipulated environment. In this sense the study also differed
from previous research on the use of online discussion forums, blogs, social network sites and
so on, many of which involve case studies of particular online environments, but all of them
involving existing online environments (e.g. Zhang and Watts, 2008; Hall and Graham, 2004;
Jansen et al., 2009; Honeycutt and Herring, 2009). The requirement of setting up and „testing‟ a
new information-sharing platform meant that the study took on some elements of a quasiexperiment, but without the availability of a suitable comparison group, only a single-group pre106
Methodology, Phase 2
test post-test design would have been possible. Although this design is commonly used to
evaluate the impacts of policies and programmes (GSR, 2007), it is considered to be
methodologically weak (Robson,1993). Consideration was given to using an interrupted timeseries design, imposing interventions throughout the course of the project, and observing any
changes following these interventions, but this was also rejected because the main objective
was to observe interactions occurring as naturally as possible.
Real-world interventions are sometimes described as „trials‟ or „field-trials‟. In the transport
literature, the impact of such trials is often studied using an experimental design – typically a
pre-test post-test design - and quantitative methods (e.g. Beale and Bonsall, 2006; Taylor and
Ampt, 2003; Taniguchi et al., 2003; Taniguchi and Fujii, 2007). Trials may be described
differently when a product, rather than a policy is being tested or evaluated. In technology
studies (e.g. human-computer interaction), prototype technologies are often „trialled‟ to find out
what happens when people use them outside the confines of the laboratory. These studies are
sometimes referred to as „field trials‟, and do not necessarily employ the language of the
experiment (e.g. Esbjörnsson et al., 2004; Axup et al., 2005, discussed in Section 2.5). This less
experimental approach to trials is more analogous with the research design in the second part of
this thesis than the policy-related trials typical within transport research, although the focus of
such studies tends to be on evaluating the technologies per se, rather than using the technology
as a context to study users' wider attitudes and behaviour.
The Phase 2 design did not, therefore, meet the criteria of a case-study, an experiment or a field
trial in their strictest senses, whilst containing elements of all three. It is argued, however, that
the research design was, in all senses but one, that of a case-study, and hence this term has
been adopted in this thesis. Whilst the information system itself was to be specially created for
the project, it would not be used in a laboratory, but in a real-world environment, and was to be
studied in relation to its impact on real-world behaviour. The research was also based on an
interpretive epistemology and employed qualitative methods, both of which are typical of the
case-study; this contrasts with the more positivist language of the social science experiment.
Moreover, it has been argued that a strong overlap exists between case-study and other
research designs, and there are particularly clear overlaps with some forms of experimentation.
Like a case-study, an experiment takes place in a specific context at a particular time, and may
be complemented by additional forms of data collection (Robson, 1993, Flyvbjerg, 2006). A
leading proponent of the case-study approach, Yin (2009) argues that case studies were long
thought, mistakenly, to be just one type of quasi-experimental design – the “one-shot post-test
only” design. Robson (1993), argues that it is possible to undertake “in-breeding between
Methodology, Phase 2
strategies” (p.169), suggesting that case-study and experimental methods might usefully be
combined in real world research:
"The inherent flexibility of the case-study, and its use of multiple methods of
investigation, make it feasible to introduce „experiment-like‟ features to develop an
understanding of the effects of the intervention." (p.169).
Having concluded that the Phase 2 design would constitute a „case-study with experiment-like
features‟, consideration was given to the development of a robust case-study design. The first
task in case-study design is to identify the unit(s) of analysis: what precisely is the „case‟ to be
studied? (Yin, 2009). The unit of analysis chosen can then have implications for how the
researcher wishes to generalise from the findings. It is possible to have one or more units of
analysis embedded within another, although the main unit of analysis is likely to be at the level
being addressed by the main study questions. As the research questions to be addressed by
this part of the PhD concerned advanced traveller information systems, and this was the area to
which analytic/theoretical generalisation was sought, the main unit of analysis was identified as
the bespoke layer of www.bristolstreets.co.uk created for the project - the Cycology website.
Following Yin‟s (2009) definition, the „case‟ - the Cycology website - represented a more abstract
concept: that of online information-sharing. Within the main unit of analysis were other,
embedded units of analysis, such as information-content and specific social-psychological
factors. The case can be placed within several wider contexts such as: commuter cycling; the
organisations where participants work or study; the city of Bristol; and travel information more
Finally, it was necessary to clarify the type of case under study, which then has implications for
the way in which analytic generalisations might be made. Yin (2009) identifies five possible
rationales for a single-case design such as Cycology: the critical case; the extreme or unique
case; the typical case; the revelatory case and the longitudinal case. Of these, the current
research constituted an “extreme or unique case” - when a case is so rare that multiple cases
cannot be studied at the same time. Analytic generalisation is possible if the case acts as an
exemplar with which to compare other cases as they arise.
Methods of data generation
As previously noted, case studies can provide a research framework in which any research
method might be used (Robson, 1993). The selected methodology for Cycology was primarily
qualitative: predominantly observation of interactions on the website, supported by participant
interviews and questionnaires to address research questions which could not be adequately
Methodology, Phase 2
answered through direct observation. This approach was informed by previous empirical studies
of online behaviour; for example, both observation and interviews were used by Hall and WidénWulff (2008) in a case-study of blogging practices among a group of students, whilst Hall and
Graham (2004) used observation, survey by questionnaire and in-depth interviews to study the
code-breaking Yahoo „e-group‟. The researcher also had administration rights to the Cycology
system, which provided access to information on date and time of log-in to the site by each
participant, and each of the markers opened. This allowed a descriptive (numerical) analysis to
be undertaken of participants‟ engagement with the website.
The combining of observation with interviews is also a familiar technique in the transport field,
although little has been undertaken so far in an online environment. One contemporary example
is the Brighton-based TWAGO project („Twitter As you Go‟) where travellers tweet or blog about
their everyday travel experiences. A classic early („offline‟) example of combining observation
with interviews was the HATS technique, developed in the 1980s by Jones et al. (1983). The
Household Activity-Travel Simulator (HATS) involved the recording of spatial and temporal data
on activity participation by household members, which then helped to structure a discussion of
the reasons for this behaviour during interviews. By recording their movements on a map and
noting the timings, respondents were, essentially, keeping an activity travel diary. The use of
travel diaries followed up with interviews is not uncommon within transport geography; for
example, Line et al. (2010) used diaries to capture participants‟ narratives of their use of ICTs
within everyday mobility, which served as a basis for in-depth interviews. These authors note
that their methods were borrowed from those commonly used within mobility studies (Line et al.,
2010). Whilst travel diaries reveal self-reported data, new technologies now allow spatial
information about trips to be recorded automatically using GPS technologies, thus creating new
opportunities for capturing observed behaviour which can then be probed during interview.
Although the Cycology project was intended to capture data about participants‟ travel behaviour
in a way not dissimilar from a travel diary, the main focus of the study was not on peoples‟ travel
behaviour per se, but on their observed use of the website. Hence, the study was perhaps
methodologically closer to contemporary research on online communities, discussed in Section
2.5 of the literature review.
Having discussed the methodological framework and use of specific methods, we now turn to an
explanation of the design and implementation of the Cycology case-study.
Selection of the case-study information system
The selected platform for hosting the case-study system was www.bristolstreets.co.uk. This
website exhibits many features of an advanced traveller information system, providing detailed,
Methodology, Phase 2
interactive information on public transport and cycling (and car driving to a lesser degree); see
Figure 7-1. Since its inception in 2007, it has hosted a variety of other types of community
information, such as houses for sale, environmental information and „quiet places‟ in the city. Its
distinctive feature is that all the information is based on a (Google) map, overlaid with different
categories of information (such as a bike layer, a bus layer, and a „quiet places‟ layer). Whilst
much of the information is retrieved automatically from a range of other information sources and
overlaid on the map, some of the layers – notably the bike layer and the quiet places layer include markers added by visitors to the site. Thus, the travel component combines formal travel
information (e.g. bus routes and times) with an interactive cycling layer on which people may
draw cycling routes and add markers, such as comments and photographs, and may also
endorse each other‟s comments. The cycling layer has been used to host a survey of cycling
facilities around Bristol – hence many comments have been posted by users with the intention of
drawing problems to the attention of the city council. The website thus provides a combination of
local formal and informal information with possibilities for social interaction, and was therefore
considered to offer an ideal web-based environment within which to build the case-study
information system for this research. Screenshots from the website are presented elsewhere in
this thesis in Figure 1-3,
Figure 8-13 and Figure 8-15.
Methodology, Phase 2
Figure 7-1: Screenshot showing “About” section of www.bristolstreets.co.uk
Methodology, Phase 2
During June and July 2009, when the observational stage of the Phase 2 empirical research
took place, total „visits‟ to the site numbered 3938 , a mean of 65 visits per day. Hits („clicks‟) on
the site over the same two months totalled 41,887, of which 8978 were on the bike layer (mean
687 total hits per day; of which 147 were on the bike layer) .
The designer of www.bristolstreets.co.uk, was sufficiently interested in the concept of testing out
new „social design features‟ to agree to build a bespoke layer of the website for the purposes of
the research; thus, the Cycology layer was born.
Population, sample and recruitment
The population from which the research sample was drawn was people who cycle, or were
considering cycling to work or study at five neighbouring organisations in Bristol. This specific
group of modal users was selected for in-depth study on the basis of findings from Phase 1 as
well as the small body of literature concerning social identity and cycling (e.g. Skinner and
Rosen, 2007, Gatersleben and Haddad 2010) and cyclists‟ preference for personalised route
information (Priedhorsky et al., 2007). Further empirical evidence of the willingness of
employees at UWE to share route information with others in the organisation was provided by a
small „bike buddying‟ scheme which was piloted in the spring and summer of 2008.
Approximately 30 existing cyclists came forward to offer route advice to other people.
The study population was limited to five neighbouring organisations because the Phase 1
findings, as well as evidence from the social psychology literature (e.g. Kramer and Brewer,
1984; Van Lange and De Dreu, 2001), suggest that people are more likely to cooperate with one
another within a defined community. At a more practical level, the sharing of cycle route
information would be more relevant to participants if they cycled to the same locality. Even
where participants in Phase 1 did not necessarily express feelings of in-group identity, they
expressed a preference for „local knowledge‟ from others within their community (in this case the
university), because these people were likely to have relevant experience and were therefore
regarded as trustworthy - this could in fact be instrumental reasoning as a proxy for issues of
social identity. The five organisations were located on the northern fringe of Bristol: UWE
A “visit” is when a particular web browser connects to the site. Figures exclude the visits and
hits by Cycology participants, which are presented separately in Chapter 8.
Figures provided by Toby Lewis.
Methodology, Phase 2
Frenchay Campus; the Higher Education Funding Council for England; the Ministry of Defence
at Abbey Wood; Hewlett Packard Labs; and a small engineering company. This allowed the
study population to be diversified beyond the Phase 1 research population, but without
compromising the project‟s local focus.
Volunteers were recruited from the five organisations using posters and flyers, bicycle user
group email lists, „snippets‟ on the organisations‟ intranet websites and the Bristol Streets
website itself. In order to try to reduce the potential bias towards experienced cyclists, people
who had recently bought a bicycle through the university‟s tax-free purchase scheme were also
approached via the personnel department (in the hope that this might attract novice cyclists). A
„snowballing‟ technique was also used by asking the selected volunteers to forward the invitation
to participate to anyone they knew who was contemplating cycling but had not yet done so (a
strategy which was not successful). An initial online questionnaire was administered to
determine volunteers‟ level of experience in cycling (generally, and specifically in this part of
Bristol) and any previous use of www.bristolstreets.co.uk or other online cycling
information/mapping etc, plus the area they would normally commute from. The aim was to
recruit people with different levels of experience as some were expected to act more as
„information-givers‟ and some as „information-seekers‟, as well as to obtain a broad geographical
coverage of participants‟ cycle routes to their places of work or study. Twenty to 25 active
participants were thought to be required for the duration of the project in order to generate
sufficient numbers and combinations of interactions, whilst keeping the group small enough to
allow a sense of social closeness within the group, as well as allowing the researcher to meet
and interview as many participants as possible. Some guidance on group size was found in the
literature, although there was little consistency and limited direct comparability with Cycology.
For example, Hall and Widén-Wulff (2008) conducted a case-study of online information-sharing
amongst a small group of only six postgraduate students. Tu and McIsaac (2002) studied
interactions amongst 51 students in an online learning environment, and found that “Many
students reported that they felt lost in the multithread discussion environment. They became
confused and frustrated, having difficulty determining “who” was talking to “whom” about “what”.”
(p143). Studies of „organic‟ online communities revolving around a hobby or interest (as
opposed to online groups within a work/study context, or those set up for research purposes)
tend to involve groups of far larger numbers, although indepth data may only be obtained from a
limited number of group members. For example, Hall and Graham (2004) studied an online
community with over 2,500 members, but surveyed 30 and interviewed eight. Taking into
consideration these factors relating to both interactivity within the group and practical matters of
data generation and management, a sample of 30 participants was selected purposively for
Methodology, Phase 2
Cycology from a group of 60 volunteers, of whom 23 eventually played an active part in the
Although one of the organisations was the university from which the Phase 1 sample had been
drawn, none of the original participants took part in Phase 2. The final sample contained people
who had been cycling to work for differing lengths of time, ranging from six months to over ten
years, and who were at that time cycling to work with different frequencies (ranging from every
day to very infrequently). Just one of the 23 participants was contemplating cycling to work but
had not yet done so. A diversity of ages and a balance in numbers between men and women
was sought. Achieving a gender balance was important, as previous research on interactions in
online communities has highlighted differences between levels of male and female participation
in online groups, as well as in the nature of their interactions (Mann and Stewart, 2000). A
summary of the sample is shown in Table 7-1 and Table 7-2.
Table 7-1: Cycology participants by age, gender and organisation
Age range
student -E
Total 6
UWE: University of the West of England; HEFCE: Higher Education Funding Council for
England; HP: Hewlett Packard; MoD: Ministry of Defence.
Table 7-2: Cycology participants – commencement and frequency of cycling to work
Commencement of cycling to
work prior to the study
years years years
Frequency of cycling to work
* One female participant is excluded from this part of the table because she had not cycled to
Methodology, Phase 2
Careful consideration was given to the type of incentives offered to participants in this project. It
was concluded that social and practical incentives may work better than financial ones for a
study of this type. For example, participants may gain personal, instrumental value from
information exchanges (e.g. knowledge of recommended routes), but also social value through
reputation-building („being an expert‟), or psychological value from supporting a project which
may also benefit other cyclists in the long run or even encourage non-cyclists to give it a try.
Careful thought was given to ways in which such benefits might be communicated in the
recruitment process and during the project.
Operation of the Cycology study
Participants were invited to attend a briefing (with lunch) in order to introduce them to the casestudy system, the research project, and to one another. The briefing was run twice in order to
maximise attendance - therefore participants were not all able to meet one another. The
researcher offered to meet those who could not attend one of the briefings on a one-to-one
basis. Thirteen participants attended the briefings, and the researcher met a further three people
one-to-one. The opportunity for participants to meet some of their fellows at the start of the
experiment was thought to be one way of encouraging both the level of interaction amongst
them during the experiment, as well as a general sense of involvement in (and commitment to)
the project. As prior knowledge of one another may have influenced their interactions during the
project, participants were asked whether they knew anyone else who had volunteered for project
(no one was aware that they did). All were given a Participant Guide and Project Information
Sheet (attached as Appendix H and Appendix I).
Participants were asked to use the Cycology website over a period of 6 weeks, starting in early
June 2009. They were invited to mark their favourite cycling route/s on the interactive map, post
comments or photographs, discuss local cycling matters, and respond to one another‟s
questions. Each marker appeared on the map as a balloon identifying the person who had
created it. Clicking on the marker revealed a comments box, to which subsequent comments
could be added, in the manner of a discussion thread. As well as the geographical markers, it
was also possible to create „floating markers‟ for comments and responses not relating to a
specific location. To personalise the messages, participants could submit a thumbnail image
such as a personal photograph or icon. In the manner of a discussion forum, participants were
sent an email digest every day, containing any markers which had been created, or comments
added to existing markers during that day. Each person had a registered password to enter the
Cycology layer of the site, which was only accessible to members of the group for the duration
of the experiment. All the interactions were observed and recorded. The researcher also had site
Methodology, Phase 2
administration rights which allowed her to analyse participants‟ browsing activity: which markers
they were looking at and when. Participants were also asked to keep a record of their activities
on the website (e.g. reasons for looking at a particular feature) and any off-line activities
associated with it, such as following up information by checking other sources, using a
suggested cycle route, or contacting another participant off-line.
The website contributions were analysed by the researcher, in order to explore within the written
text any discourses of trust, social identity or pro-social behaviour, as well as seeking evidence
of any behavioural change (e.g. trying out new routes). However, it was necessary to go beyond
an interpretation of observed behaviour in order to achieve a better understanding of the socialpsychological factors under study, particularly those concerning motivations for sharing
information and reasons for trusting others (or not). These factors were initially explored through
a short questionnaire, then probed through one-to-one interviews at the end of the experiment;
the interviews also provided an opportunity to explore participants‟ reflections on their
experience of using the site, any influence on attitudes, intentions and behaviour, and possible
suggestions for improvements to the social features of the site.
Ethical considerations were also given due consideration in the design of the study. In addition
to the following of standard university procedures for ethical social research, concerning
confidentiality, anonymity and the right to withdraw (all participants signed a consent form
confirming their understanding of these issues), thought was also given to additional ethical
issues relating to the use of an online discussion forum. This included the adoption of a code of
conduct for „appropriate use‟ of the website, adapted from those used by the Students‟ Union
and Yahoo for the online forums which they host (attached as Appendix J– Code of conduct for
on-line discussions). Participants were informed of the option of using a pseudonym on the
website, and were advised not to reveal personal information such as home addresses.
Content of questionnaires and interviews
Both the questionnaires and interviews were designed to complement the observational data by
giving participants an opportunity to explain their own contributions, their reactions to those of
others, and broader views about the website, both as they had experienced it and what they
thought its wider potential might be. The questionnaire and interview questions were designed to
answer the detailed Phase 2 research questions, in particular research questions 4.3, 4.4 and 5
concerning, respectively: behavioural and attitudinal effects; social-psychological processes; and
wider potential applications in the field of advanced traveller information.
Methodology, Phase 2
Firstly, the short questionnaire was designed, comprising 13 open questions (attached as
Appendix F). The main purpose of the questionnaire was to capture participants‟ immediate
views on their experience of the Cycology project as soon as the six weeks had passed,
including: their evaluation of the information posted; their motivations for contributing (or not);
any perceived changes to their own attitudes and (travel) behaviour; and thoughts about how
such a website might work in different contexts. This was in anticipation of a time lag between
the end of project and the earliest time when interviews could be held, the main summer holiday
period having now started. The questionnaire also served a useful function in helping the
researcher to prepare for each interview and refine the interview guide for each individual, based
on the answers given in the questionnaire. Thus, issues raised in short statements in the
questionnaire could be probed and full explanations sought during the interviews. The interview
guide template is included as Appendix G. All interviews were transcribed for analysis within
Data analysis
A thematic analysis of the website contributions, questionnaire responses and interview
transcripts was undertaken according to the same procedures and with the same philosophical
underpinning as described in Section 4.3. A list of NVivo codes is attached as Appendix L. Three
further forms of analysis were also undertaken: a descriptive numerical analysis of levels of
activity on the website using Microsoft Excel; the establishment of a „typology‟ of Cycology
participants, and non-cross sectional (holistic) analysis of individual participants. The latter two
techniques are explained below.
Establishing typologies
Typologies are forms of classification which help to explain the segmentation of the social world
or specific ways in which phenomena can be characterised or differentiated (Ritchie et al. 2003).
This involves creating multi-dimensional or multifactorial classifications in which categories are
discrete and independent of one another (an individual feature can only be assigned to one
category). Typologies were thought to be useful for the Phase 2 analysis, as they might, for
example, provide findings of interest to developers of traveller information systems incorporating
user- generated content, by indicating the different types of people needed to make an online
group (or „community‟) of travellers function well.
Kozinets (2010) proposes a typology of interest to the current research, which categorises types
of interaction within online communities. "Participation can move from a factual and
informational type of exchange to one that effortlessly mixes factual information and social, or
Methodology, Phase 2
relational, information." (p34-35). In cruising communities, participants‟ interest in the
consumption activity is low, as are social relationships with other members. A bonding
community is characterised by deep and long lasting relationships amongst members, but a
weaker focus on the unifying consumption activity (for example, a social networking site). The
primary purpose of a bonding community is to meet participants‟ relational needs. Another type
of community is where the focus is very much on the consumption activity, and less on social
bonding. This type of community is likely to involve the sharing of information, news, stories and
techniques about a particular activity. Kozinets terms this type of community a geeking
community. Many newsgroups and blogs would fit into this last category, where members share
detailed information but do not engage deeply in social relationships. Kozinets (2010) maintains
that the modes of interaction in these communities are predominantly informational. The fourth
type of community combines both strong social bonds with detailed information about the
unifying interest. These are termed building communities, and often emerge from website
forums and virtual worlds. Here, the interaction is both informational and relational. It was of
particular interest to see whether the Cycology group would constitute a geeking community, or
perhaps develop into a bonding or a building community (if indeed it could accurately be defined
as an „online community‟ at all – a matter to be discussed in Section 11.4).
One way in which the establishment of a typology was thought to be useful in the analysis of
Phase 2 data was to inform the selection of a small number of cases for holistic (non-crosssectional) analysis.
Non-cross-sectional analysis
Non cross-sectional analysis may also be termed contextual, case-study or holistic analysis
(Mason 2002). These forms of data organisation focus on discrete parts of the data set, with the
aim of documenting something about these parts in particular (e.g. observational, questionnaire
and interview data relating to a specific individual). Mason (2002) suggests that “it is a practice
guided by a search both for the particular in context rather than the common and consistent, and
the holistic rather than the cross-sectional” (p.165). Two reasons for taking a non-cross-sectional
approach to the Phase 2 analysis were: a) a desire to gain a sense of the distinctiveness of
different elements of the dataset (i.e. participants as individuals); and b) a wish to use this
method in addition to cross-sectional indexing, in order to build explanations based on two
alternative ways of „slicing‟ the dataset.
The analytic logic for this approach is that explanations are derived from the analysis of, or
comparisons within, cases or „wholes‟. Instead of comparing specific features of one person's
account with similar features of another person's account, as if they were like-for-like and free of
Methodology, Phase 2
context, the aim is to compare the overall explanation of a phenomenon within one person‟s
account, with the explanation of the same phenomenon in another person's account. For
example, with regard to Cycology, this could relate to an individual's propensity to contribute to
the website within the context of their personality, usual behaviours, likes and dislikes, and time
commitments. The mixing of both cross-sectional and non-cross-sectional data analysis is a
method explored by Freudendal-Pedersen et al. (2010), who use the terms horizontal (thematic)
and vertical (holistic) analysis in relation to a study of the ways in which individuals narrate their
everyday life-related choices, and the unintended consequence of these choices on their
mobility patterns. A vertical analysis was carried out of individual participants in the Cycology
project, using all data sources pertaining to each individual to complement the main horizontal
(thematic) analysis, in order to bring out individual voices from within the data. Four such vertical
analyses are summarised in Section 11.4.2, each representing a „type‟ within the typology which
was developed.
7.10 Chapter Summary
This chapter has presented the methodology used for the second empirical research phase
within the thesis. Starting with the context of online cycling information, it was explained that no
„naturally occurring‟ information system had been identified in which social interactions were of
the quality required to answer the research questions, and therefore a bespoke system needed
to be developed, incorporating design features which might stimulate social interaction. This took
the form of a specially designed layer of www.bristolstreets.co.uk, entitled Cycology. The
research design was one of a case-study „with experimental features‟ (as Cycology had been
developed for the project and hence did not constitute a „naturally occurring phenomenon‟ in the
strictest sense of the case-study approach). A mixed method approach to data generation was
employed (observation, questionnaires and interviews). Thirty people, of whom 23 eventually
took part, were selected purposively to use the Cycology system over six weeks. Data were
analysed using thematic analysis, descriptive numerical analysis, typologies and non-crosssectional analysis. The findings of the research will now be reported in the following four
Findings, Phase 2 – Descriptive analysis
Chapter 8 : Findings, Phase 2 - Descriptive analysis of the
Cycology website
As discussed in earlier chapters, the Phase 1 findings had provided new knowledge about the
context of informal travel information and word-of-mouth processes (who tells what to whom,
and when), its influences on attitudes, intentions and behaviours, and underlying socialpsychological mechanisms. In particular, Phase 1 had delivered new insights into the role of
word-of-mouth as a channel for „background‟ normative influence, affecting general beliefs and
attitudes about travel within which the decision process for specific trips is later framed, as well
as social factors found to be associated with informal information in the context of active trip
planning (depicted, respectively, in the top and bottom left sections of Figure 6-1, p95). Neither
of these areas had been discussed in the existing literature on traveller information (within
transport studies), as this has conventionally addressed the use of information only within an
active decision process and from an individual, not a social, perspective.
Phase 2 of the research had then set out to explore informal information and word-of-mouth
processes within an applied context – that of advanced traveller information (research questions
4 and 5). The tasks in Phase 2 were to:
validate the earlier, general findings in the applied context of a real-world traveller
information system;
explore the operation of the identified social mechanisms in a small-group context;
study the identified social factors in greater depth.
The findings reported in this and the next two chapters address the fourth main research
question of this thesis:
4. How do word-of-mouth processes and informal information-sharing operate within an
emerging web-based system incorporating user-generated travel information?
This question was divided into four sub-questions corresponding with the main areas of Phase 1
4.1 For what types of informal travel (cycling) information is the case-study system used?
4.2 What word-of-mouth processes (e.g. patterns of interaction) can be observed within this
small group context?
Findings, Phase 2 – Descriptive analysis
4.3 What influence (if any) has word-of- mouth information exerted on participants‟ selfreported attitudes, intentions and travel behaviour within the case- study?
4.4 How do social-psychological mechanisms appear to function in this online context?
Although the data generated through all three methods used in Phase 2 (observation,
questionnaire and interview) were integrated to answer all four sub-questions, the answers to
the first two - reported in this chapter - drew principally on observational data from the Cycology
website, whilst the second two, reported in Chapter 9 and Chapter 10, drew mainly on the
interviews and questionnaires. A benefit of the mixed-method approach was that participant
accounts could help provide a better understanding of some of the observed activity on the
website, as well as elucidate people‟s own experience of using it and the ways they believed it
had influenced them. Conversely, observation of actual activity on the website provided a means
of validating or questioning participants‟ own accounts. This also meant that the analysis could
be undertaken using both the group as a whole, and each individual, as units of analysis. Thus,
in Sections 8.1 and 8.2 below, analysis of website use and patterns of content over time are
presented in aggregate (use by the whole group). In Sections 8.3 and 8.4, data on types and
patterns of participant involvement are disaggregated into individual users . The psychological
and behavioural findings reported in Chapter 9 and Chapter 10 emerged principally from a
thematic qualitative analysis, in which themes were identified horizontally across the whole
group, but illustrated with individual examples.
This chapter describes the findings of a descriptive numerical analysis and a qualitative content
analysis of messages posted on the website, the viewing of posts, levels of engagement by
different participants, patterns of interaction, and culminates in an analysis of the two „most read‟
discussion threads. Furthermore, by providing a comprehensive description of the activity on the
website, it sets the context for the subsequent two chapters on attitudinal and behavioural
influences, and underlying social-psychological processes associated with information-sharing
(Chapter 9 and Chapter 10, addressing research questions 4.3 and 4.4 respectively). The
descriptive analysis also supports Chapter 11 of this thesis, which considers implications for
traveller information systems more broadly.
Each participant was given a pseudonym for the reporting of the findings, which will be used in
this and the remaining Chapters.
Findings, Phase 2 – Descriptive analysis
Following a numerical overview of activity on the website, this chapter addresses sub-questions
4.1 and 4.2. in the following sub-sections:
4.1 For what types of informal travel (cycling)
8.2 Content of website posts
information is the case-study system
8.4 Detailed analysis of two interactions
4.2 What word-of-mouth processes (e.g.
8.3 Levels of participant involvement
patterns of interaction) can be observed
within this small group context?
Numerical overview
Over the six week period of the Cycology project, 132 postings were added to the site by the 23
participants, of which 77 elaborated on routes drawn on the map, and 55 comprised general
comments, questions or responses. Eighty postings formed part of 29 short discussion threads,
the number of comments in a thread ranging from 2 to 5. The 13 female participants contributed
65% of the postings (mean number of postings by women = 6.6), whilst the 10 male participants
contributed 35% (mean number of postings by men = 4.6). The overall mean number of posts
was 5.8. The lowest number of contributions was 1, and the highest 14. Four (female)
participants accounted for 34% of the postings. Figure 8-1 provides a screenshot of the website
at the end of the first week of the project. Figure 8-2 shows that 46% of contributions were made
during this first week, as participants enthusiastically drew their routes on the map, after which
the subject matter widened but contributions became less frequent. Figure 8-2 also shows that
at least one message was posted on 26 days out of the 30 working days (i.e. Monday to Friday)
of the project. Interestingly, participants were not prevented from posting at the weekend, but
never did, demonstrating that the website tended to be associated with commuting and work –
an observation supported by people‟s accounts in the interviews. Some of the most active
contributors commented that checking the website was part of their daily routine, and something
they might do when they first arrived at their desks, or when they needed a short break.
Findings, Phase 2 – Descriptive analysis
Figure 8-1: Screenshot of the Cycology website at the end of week 1.
Findings, Phase 2 – Descriptive analysis
Viewings of markers on the website numbered 1059 over the six weeks; the breakdown over
time is depicted in Figure 8-3. 89% of these markers were posts, or groups of posts (discussion
threads) created by participants, the rest being markers already present on the website, such as
those showing the location of cycle shops or cycle stands. The number of markers opened per
individual ranged from 1 to 127 (mean = 46). There was less of a gender imbalance than was
the case for the number of messages written, with women representing three of the five top
„readers‟ (whereas the five top „writers‟ were all women). The mean number of markers was
similar for men and women (45 and 47 respectively). A comparison of male and female
participation was of interest because previous research on interactions in online communities
has highlighted differences between levels of male and female participation in online groups
(men tending to participate more), as well as in the nature of their interactions: for example,
Mann and Stewart (2000) suggest that men tend to engage in “report talk” (focussing on
instrumental information) and women in “rapport talk” (placing greater emphasis on relationshipbuilding). However, these findings were not supported by the present research, as female
participants contributed more than male participants on average, but the content of their posts
did not, overall, differ in a qualitative way from those of male contributors (male participants were
not necessarily more „factual‟, nor female participants more „conversational‟). As was the case in
Phase 1, gender differences were neither clear nor consistent across this small sample.
Figure 8-3 shows that the number of markers being opened decreased over time, following the
same pattern as the number of posts written. Each of the 132 markers was opened eight times
on average. However, these figures must be qualified with the important observation that it was
not possible to know how many participants read the posts in the daily email digest without
following the links to the website. Several participants noted at interview that they were happy to
read the email digest unless they wanted to look at a specific route on the map, and this was
reported to have happened increasingly with time, reflecting the tendency perhaps for the topic
of the posts to broaden out from routes, and therefore becoming less dependent on the map.
Findings, Phase 2 – Descriptive analysis
Figure 8-2: Number of messages posted over time
Number of messages posted over time
number of messages posted
Figure 8-3: Number of markers viewed over time
Number of markers viewed over time
number of markers viewed
Findings, Phase 2 – Descriptive analysis
How might the level of activity on the Cycology website be assessed overall? There were no
specific expectations, as no similar „real world‟ website had been identified with which to
compare it: both the research design and the website itself were experimental. However, the
Bristol Cycling Campaign discussion forum was being observed to provide a loose comparison,
and the number of messages posted to these groups per day averaged 4.9 during the same
period in June/July 2009. This is not greatly different from the average number of posts to the
Cycology site per working day (4.4), despite the much larger membership of the cycling
campaign discussion group - comprising just under 200 members. During the same time period,
the cycle layer of the main bristolstreets website was attracting a lower level of activity:
approximately 1.4 posts per working day (as most markers were added to the main site as part
of a survey about cycling facilities for the City Council, over two thirds identified hazards and
locations for suggested cycle lanes; only 8% constituted routes). However, in the later weeks of
the Cycology project, the number of posts per day rarely exceeded three, and this number
seemed unlikely to rise without the „injection‟ of new members to add new routes and discussion
topics. Some participants expressed, at interview, a disappointment in the number of posts and
level of interaction, although this was affected by the extent to which an individual took part in
discussion threads (a matter returned to in Section 8.3). However, others reported that the level
of interaction was sufficient to maintain their interest, and most said they found the project useful
and/or enjoyable. In terms of how well the website functioned as a „product‟ which might be
further developed (not a research question as such, but an area of interest in relation to potential
applications in the field of advanced traveller information), one area which did not „take off‟ was
the posting of routes other than commuter routes, such as leisure routes, but this was attributed
to the strong commuting focus of the project. A „product tester‟ might have been disappointed
that as many as one third of the participants posted only twice (and one only once), and one
third of the total posts were written by only four participants, but this does concur with literature
on participation in online communities, which suggests that the majority of posts tend to be
provided by a small minority of the participants (e.g. Ling et al., 2005; Zhang and Watts, 2008).
Mean number of posts per day was calculated by dividing by five, as the website was not used
at weekends.
Findings, Phase 2 – Descriptive analysis
Content of website posts
This section presents a numerical summary of website content and trends in content over the six
weeks of the project, and a thematic, qualitative analysis of website posts. Together with the
previous section, it answers sub-question 4.1:
4.1 For what types of informal travel (cycling) information is the case-study system used?
As would be expected, given the focus on commuter cycling and the recruitment within places of
work or study, the major topic of discussion on the Cycology site was cycle routes between the
participants‟ areas of residence and their places of work, drawn on to the map and annotated
with comments, some short and functional, and others more descriptive (examples will be given
below). Some other routes around the city were added, such as cycle routes to railway stations,
between residential areas, and to the popular Bristol to Bath cycle path. Some comments about
leisure cycling on the Bristol to Bath cycle path were also made, but beyond that, little discussion
about leisure routes occurred. Apart from the description and discussion of routes, postings
generally took the form of up-to-date warnings, such as an incidences of bike theft or broken
glass on a roundabout, discussion of general cycling issues such as taking bikes on trains or the
Bristol Cycling City initiative, notification of a cycling event (the Big Bike Breakfast), and musings
and observations about the cycling experience in general or specific occurrences on the journey
to or from work. A list of all posts and threads is attached as Appendix M.
As well as providing their own information and observations, most participants also posted
questions, which often led to short discussion threads. However, some questions also went
unanswered, the questionnaires and interviews revealing this to be a source of disappointment
to some. A significant reason for this was the limited number of participants; 23 people did not
appear to be a high enough number to be able to answer queries about cycle routes from all
parts of the city since participants did not, between them, have knowledge of cycle routes in all
areas. This suggests that greater numbers would be required if an information-sharing forum
such as this were to be sustainable in the longer term, although the advantages of greater
breadth must be weighed against the possible loss of „intimacy‟ as group size increases (see
Section 10.3 for a full discussion of group size effects).
Numerical analysis
A numerical summary of the content and function of all posts was provided by placing each
within one of seven categories, which are presented in Figure 8-4. The first three categories are
grouped together as those pertaining to routes. The largest category (35 posts) comprises
Findings, Phase 2 – Descriptive analysis
comments tied to routes drawn on the map, and a further 19 provided detail of features along the
route. These posts were all proactive (i.e. offering information without it being requested) and
peaked in the first week of the project.
Figure 8-4: Number of posts by type
Twenty three messages were posted in response to proactive route comments, and this
sometimes included feedback when someone had tried out a route and then offered views on it.
These responses followed the same pattern over time as the proactive route information. The
second group of categories comprise direct questions and responses to them. There were
slightly more questions than answers, and a slight time lag between the two, with most questions
in the first and third week and most responses in the fifth week. The third group of categories
(general comments and responses) comprised proactive comments concerning non-route
related matters, plus a small number of responses to them. These comments peaked in the
fourth week of the project. Many participants indicated in the interviews that by this stage they
had said most of what they had to say about their routes, and started to widen the subject
matter. However, the number of general posts and responses per week was relatively low
(between 2 and 7). Charts showing the distribution of messages over time in each of the three
groups of message categories are provided in Appendix K. Figure 8-5 shows that when the
numbers of all route-related comments are combined, and compared with the combined
numbers of questions, answers, general comments and responses, over half of route-related
Findings, Phase 2 – Descriptive analysis
messages were posted within the first week, whilst questions and general comments were more
constant over the 6 weeks, albeit with a reduction towards the end. In total, route-related posts
constituted a slight majority of all posts (57%), and, as has been previously noted, some of the
questions also related to routes. This gives rise to speculation about how the pattern of
messages would have continued had the project gone on for longer. Unless more „news-worthy‟
incidents were to occur along the routes of the existing participants, it seems likely that more
people would have been required to join in order to bring fresh contributions.
Figure 8-5: All route-related comments compared with all other comments over time
All route-related comments compared with all
other comments over time
all route related
all other
8-14 June 15-21
29 June 5 July
6 - 12
13 -17
Qualitative analysis
The previous section showed trends in content over time, by grouping posts into broad
categories according to subject matter (route or non-route related) and function („pro-active‟
statement or question; „reactive‟ response). In this section more detailed attention is given to the
content of the posts, drawing out themes which run across these broad categories, as well as
the tone in which they were written. The main themes identified are wayfaring (geographical
route descriptions), safety, cycling infrastructure, sharing space with other road/path users, and
sensory, aesthetic and affective descriptions of the cycling experience. Many posts fell into more
than one of these categories. The titles of posts and groups of posts (threads), categorised
according to these themes, are attached as Appendix M.
Across all the themes, many postings combined both instrumental/geographical information,
such as steepness (the words hill, hilly or steep appeared some 45 times), traffic volumes and
infrastructure features such as quality of road surfaces, with subjective descriptions of the „lived‟
Findings, Phase 2 – Descriptive analysis
experience of cycling the route. According to the classification of formal and informal information
used in this thesis, instrumental/geographical (factual) information may be categorised as formal
(obtained through word-of-mouth rather than official sources), whereas more personal,
subjective evaluations are categorised as informal. Therefore, both informal and formal
information were communicated through word-of-mouth on the website, and many individual
posts contained a combination of both. This means that although the research question
addressed by this analysis concerned informal information, the findings actually refer to both
formal and informal information delivered through word-of-mouth.
A small minority of the wayfaring comments were short and functional, in the vein of:
Daily trip to work – mainly off road, bit hilly. (Kate)
This is my normal route to work. I go along Gloucester Road and Filton Avenue
to avoid losing height. (Henry)
However, the following represents a more typical route comment, combining factual information
(visibility, speed humps) with evaluative and experiential commentary:
Good route. Haven't had many problems with aggressive drivers. The roads
don't have cars parking alongside, so there's lots of visibility. The only
problem is coming down Braydon Avenue. The road surface is terrible and
when combined with the speed humps, you spend a lot of time in the middle
of the road. But the road is so bad, I've not had trouble with cars wanting to
go too fast. (Jess)
This combining of the „factual‟ with the „evaluative‟ is in line with the findings from Phase 1 about
the way in which formal and informal information are thought to complement one another,
although in the Phase 1 interviews, „formal‟ information was generally associated with
information from official sources (e.g. printed timetables) rather than factual information received
through word-of-mouth. Furthermore, in Phase 1 the content of information which interviewees
had reported obtaining or offering through word-of-mouth had been categorised in the following
ways: options and ideas; specific factual detail; or subjective advice. In Phase 2, specific factual
detail was usually combined with subjective advice, but general options and ideas (which in
Verdana font is used to distinguish verbatim comments on the website from participant quotes
arising from interviews and questionnaires (italicised, Arial font).
Findings, Phase 2 – Descriptive analysis
Phase 1 had referred to matters such as suggesting alternative transport modes) were absent,
presumably because the information offered and requested was specific to a narrow group of
users travelling to and from similar places, and using the same transport mode.
Other route comments combined factual route information with more whimsical observations. In
this example, a route plotted on the map was supplemented with a remark about a cat :
A nice route (there is a black cat on a wall who appreciates a tickle behind the
ears) but bendy so you can't go fast, there are too many people walking dogs
and toddlers on it. (Elaine)
Another participant remarked in her interview how much she had enjoyed this comment, and had
looked for (and found) the black cat when she cycled along this route – hence, although she
was already familiar with the route, she believed that the additional commentary had improved
her experience of cycling along it (a matter to be discussed in Section 9.4 on changing the
experience of the commute). Hence, such descriptive comments can make journeys more
interesting, which may increase positive attitudes towards it – this will be discussed in Section
The safety theme was manifested mainly through the comments about avoiding or dealing with
motor traffic, which appeared in some twenty of the postings, and concerns about using unlit
paths at night. A general consensus about preferring quiet routes was apparent, and there were
some specific discussions about how to deal with traffic at difficult junctions. Safety concerns
with unlit cycle paths provided another common area of concern. A sense of fear was sometimes
communicated within comments about road traffic or unlit paths:
I guess that the cycle path is dark and scary in winter. (Sally)
Grievances about road infrastructure which did not favour cyclists were also shared amongst
participants who had experienced similar problems. For example, a comment that a traffic island
designed for cyclists and pedestrians was too narrow provoked an emphatic response of :
”totally totally agree”. Other comments about infrastructure covered topics such as a poor
cycle lane, inadequate cycle access to the university, and poor road surfaces.
Matters of shared space with road or path users other than motorists were also common topics.
The general tone of comments about drivers‟ behaviour towards cyclists was uncomplimentary
or sarcastic, although not aggressive, for example:
Findings, Phase 2 – Descriptive analysis
Note: i've been screamed at by motorists a few times along Kellaway Ave.
No doubt for various injustices i have caused them. (Andy)
After cars, pedestrians were the most commonly cited annoyance, evidenced, for example by a
thread entitled “Gripe with pedestrians”. Although conflicts of interest were apparent, issues
with adults, children and dogs were often conciliatory in tone:
To be fair though, i think i tend to do the same when on foot. (Doug)
Smiling and giving way avoids confrontation. (John)
There were occasional warnings of more unusual obstacles. One contributor noted within her
description of a route through a park: watch out for the hissing geese! (Sarah). Another
had experienced an alarming but ultimately edifying encounter.
A deer jumped out at me about here on the cycle path behind HP from the
woods. I managed to skid to a halt without hitting it and then it bombed off
down the path. Be warned! However, it did make a nice addition to an
otherwise soaked journey and I've definitely never been so close to one
before. (Rick)
Moving towards the more sensory and aesthetic aspects of the cycling experience, many
participants commented on the scenic aspects of particular routes, often communicating a real
sense of pleasure in their journeys. Some postings also referred to the sensory pleasure of
cycling at speed downhill:
Sometimes i go home slightly different so that I can go down that big hill in
Lockleaze (yippie). (Andy)
The enjoyment of the journey is conveyed through this part of a discussion thread:
This is good if you have a mountain bike or it's not muddy! you do have to
carry your bike up a flight of steps first though to get up on the hill then it's
field tracks, nice and quiet and good views so for me worth the extra effort.
….I've got onto the track from a lane / car park off Shaldon Road before and
it is lovely - good for dog walking too + in spring, bluebells in the woods!
Findings, Phase 2 – Descriptive analysis
Finally, participants occasionally contributed more general thoughts about the cycling
experience. In the following example, a sense of empathy with other cyclists and the positive
feelings this engendered were clearly expressed:
So my cycle home yesterday started off as one of my most unpleasant to
date. It was absolutely throwing it down. It was nice to see I was joined by
so many other cyclists- also soaked through. The best part was that everyone
was so happy and friendly! I spoke to more people on the way home then I
have done as any other time on a bike. By the time I got home I was in a
better mood then I'd been in all afternoon, and then the sun came out :) I
guess the shared "ordeal" brought out the friendliness in everyone. Thought
I'd share that good cycling-in-the-rain experience! (Alice)
It became clear in the interviews that this type of informal comment had played a role, for some
people, in reinforcing positive attitudes towards cycling, and helping to create a sense of group
identification – matters to be discussed in Sections 9.2 and 10.2 respectively.
The over-riding tone of the interactions was friendly: a feature mentioned by many of the
participants in their feedback questionnaires and interviews. Some participants and discussion
threads were particularly „chatty‟, others more factual. Just one of the 23 participants made
comments which appeared brusque and caused slight offence to one or two others. Three
participants, one female, two male, who lived in the same area (but did not know each other)
quickly established an informal, conversational style, swapping experiences of particular routes
and addressing one another by name. Other interactions were more „functional‟ in tone, such as
a discussion among four male participants about what one was „supposed to do‟ when
negotiating a difficult junction. Some contributions were humorous, and some participants
(mainly female) often adopted a self-deprecating style in terms of their cycling competence. This
is consistent with Mann and Stewart‟s (2000) suggestion that men engage more in “report talk”
and women in “rapport talk”, although, as previously noted, this was not a difference which could
be observed to any significant degree.
Overall, providing feedback on the project, one participant summed up the content of
contributions as follows:
“There seemed to be a mix of submissions ranging from the highly informative and quite
serious to some more general (or fun) comments more about the experience of cycling in
Bristol in general. Both types were interesting and complemented each other well,
meaning that the site was very technically useful, but also not too dry and boring”.(Adam)
Findings, Phase 2 – Descriptive analysis
Another participant remarked that the mix of content made it feel “more like a conversation
between people than a formal exchange of data”, which highlights the social nature of the
information and process whereby it was shared.
Not all participants were so positive about the information content, some saying they found some
content to be trivial or mundane. Indeed, some active contributors remarked that they
considered some of their own comments to be trivial but were keen to keep the interactions
going (these participants were perhaps exhibiting pro-social behaviour and were conscious of
wanting the research experiment to be a „success‟, an issue to be discussed in Sections 10.5 pro-social behaviour, and 10.7 - reflections on the methodology).
Levels of participant involvement
This section addresses sub-question 4.2:
4.2 What word-of-mouth processes - e.g. patterns of interaction - can be observed within this
small group context?
In the figures below, data are disaggregated into individual participants.
Figure 8-6 compares the number of posts written by each participant as a percentage of total
posts, with the markers opened by each participant as a percentage of the total markers opened.
This scatter graph shows that the two activities were roughly in proportion for most participants;
for example, the person writing the most also read the most, and the person writing the least
also read the least (although, as previously mentioned, some people who rarely went to the
website stated in interview that they had nonetheless continued to read posts within the email
digest). The closer an individual sits to the 45 degree line, the more their share of the posts
written was in proportion to their share of the posts read. This is of little surprise, as someone
logging onto the site to post a message might well also take the opportunity to read the other
posts already there. Unless they wished to see a route on the map, those who simply wanted to
read a message and not post themselves could do so by reading the email digest. However,
there were some discrepancies, i.e. people whose share of the markers read was higher than
their share of the messages written (e.g. John, Rick, Julie), and, especially notable, a group who
wrote a greater proportion than they read (e.g. Elaine, Marion). The previous caveat still holds
for the latter group - they may have read a great deal more but via the email digest rather than
by going to the website and clicking on markers. However, the chart does give an indication of
participants‟ general level of engagement with the website, and the extent to which they
engaged both as contributors and readers. Those furthest below the line might be categorised
Findings, Phase 2 – Descriptive analysis
as „information receivers‟, whereas those furthest above it might be classed as „information
givers‟. For those closer to the line, their „information-giving‟ and „information-receiving‟ activities
were more in proportion, with those to the top right of the chart being the more active in both
domains, and those to the bottom left being more inactive in both areas. In terms of participant
attributes, the five most frequent contributors to the site were all female, and four of the five had
begun cycling to work within the previous two years (two had begun within the past six months).
Figure 8-6: Writing posts and opening markers, by individual
posts written by each individual as % of total
"information givers"
Mike Adam
Henry Ben
"information receivers"
Phil, Jim, Andy,
Helen, Jess, Sarah
markers viewed by each individual as % of total
One dimension which this chart fails to reveal is those people who participated little in terms of
observed website activity, but were nontheless enthusiastic about the project in their interviews
and questionnaires, stating that they had read the daily email digest with interest and had
formed opinions on some of the posts (for example Jess, Phil and Esther). Conversely, some
participants with high levels of website participation declared themselves to be less enthusiatic
during interviews than the graphs might lead one to expect (e.g. John and Elaine).
Findings, Phase 2 – Descriptive analysis
Level and patterns of interaction
A broad indication of the level of interaction on the Cycology is provided by the number of posts
which appeared in the short discussion threads. Eighty posts (58%) formed part of 29 threads,
and most participants reported that they were satisfied with the level of interaction, although
some were disappointed with it. Many thought that although the level of interaction was
satisfactory, it could also have been improved. As previously noted, not all questions were
answered, and the length of threads (between 2 and 5 messages) was rather short compared
with more „traditional‟ email discussion forums such as those of the Bristol and Oxford Cycling
Campaigns. This is likely to be a reflection of the different content of the Cycology threads (i.e.
mainly route-related) compared with the two cycling forums, in which discussion topics are more
diverse and often rather controversial. Moreover, the obvious point that the Cycology group was
much smaller than the membership of the other forums also meant that the reservoir of
knowledge (and possibly range of opinion) was narrower.
In this section, the data are once again disaggregated to analyse levels of interactivity by
individual, showing how many threads each took part in, who was more reactive and who more
proactive, and how many times each person‟s posts were read by others (in total). Network
diagrams are also provided to give an overview of the strongest connections (whose markers
each participant opened most), and the connections within two geographical sub-groups.
Figure 8-7 shows the number of times each participant started or joined a discussion thread
compared with the number of times their posts did not elicit a response. This is of interest
because it may have affected participants‟ evaluation of the level of interaction on the website.
For example, the participant who appeared at interview to be the least satisfied with the amount
of interaction responded to only two messages and received no response to six of his own posts.
Conversely, the people who were involved in the most interactions (Alice, Laura, Sally, Rachel
and Julie) all expressed relatively high levels of satisfaction with the project.
Findings, Phase 2 – Descriptive analysis
Figure 8-7: Number of posts in „threads‟ and number of stand-alone posts written by each
Number of posts in "threads" and number of stand-alone posts
written by each participant
Number of posts in threads
Number of stand-alone posts
How far is it possible to gauge from the data who the most popular contributors were? Figure
8-8 indicates that the relationship between writing and being read was not as clear as one might
expect. The person who wrote the most (Alice) was not the person whose posts appeared to be
read the most (this was Marion). This might to some degree reflect the perceived quality (or
relevance, or interest) of the content. In fact, several other people, during interview, mentioned
Marion‟s posts as examples of the type of comment which they liked. However, care must be
taken in drawing too many conclusions from these figures, as other factors, unrelated to the
content of the posts, might have influenced the frequency with which they were read. For
example, posts written at the beginning of the project, when enthusiasm was at its highest, might
have been more likely to be read on the website than those posted later, which might have been
more frequently read in the email digest. In fact, the six most popular markers were created in
the first two days of the project, and all constituted discussion threads. Moreover, comments tied
to routes were more likely to be read on the website (and hence this was recorded by the
system) than non-route related comments, because routes needed to be seen on the map,
whereas, non-route comments could just as easily be read on the email digest (hence this was
not recorded).
Findings, Phase 2 – Descriptive analysis
Figure 8-8: Posts written per person compared with number of times their posts were
read by others
Percentage of total
Posts written per person compared with number of times their
posts were read by others
posts written by each individual as % of total
Number of times each participant's posts were read by others, as % of total
These data were also affected by the fact that responses to an existing comment retained the
same title as the original comment, and were automatically recorded by the system as having
been written by the person who wrote the original post, not those who wrote the subsequent
responses. Therefore, a person who started several threads would be attributed many more
„hits‟ than a person who tended to respond to others‟ comments. The popularity of Marion‟s
contributions therefore partly reflects the fact that she started four discussion threads, whilst
Alice started only two, despite being the most frequent contributor and adding to four threads
started by others. It might be suggested that the popularity of the participants with the most „hits‟
was not entirely undeserved, however. The fact that their comments began discussion threads
could mean that their posts were intrinsically more interesting than those which did not attract a
response, or were simply relevant to more people because of the area where they lived.
Finally in this section, an analysis was carried out of who looked at whose posts (i.e. whom the
markers they opened had been created by). A matrix was created of the markers opened by
each individual and converted into a network diagram: Figure 8-9. To prevent the diagram from
becoming too complex, a connection is only shown between two participants where one person
opened markers created by the other five or more times (thus, an arrow pointing from „person A‟
to „person B‟ shows that person A opened markers by person B at least five times) . Two-way
Findings, Phase 2 – Descriptive analysis
arrows show where this was reciprocated by the other person. The size of the circles give an
indication of the overall number of times each person‟s markers were opened by the whole
group (hence Marion has the largest circle).
Figure 8-9: Who opened whose markers – whole project
Despite omitting a large number of smaller connections (fewer than 5 markers opened), and
three people who neither opened five markers of any one person nor whose own markers were
opened five times by any one person, the diagram provides an impression of the complexity of
the interconnections, and the lack of obvious patterns of interconnectivity. Even the two „outliers‟
on the bottom left (Sarah and Helen) were actually connected to several others, but opened
other people‟s markers fewer than five times. Moreover, it is notable how few of these „strong‟
links were reciprocal. However, a more detailed analysis, involving all the connections, revealed
a tendency for markers to be opened slightly more by people who lived in the same area, with
clusters of people - often those taking similar routes - tending to look at one another‟s posts and
Findings, Phase 2 – Descriptive analysis
to respond to them. Hence, there was a small cluster of people living south west of the five
workplaces who tended to interact with one another more than with others and to look at each
others‟ markers more frequently; another such group living north west, and another living east of
the employment area. The interviews and questionnaires revealed that most people were most
interested in the comments which they described as “useful”, which was often synonymous with
being geographically relevant (although they could also be „useful‟ in terms of serving a more
social purpose, as will be discussed in a later section). The website data confirmed that most
people were most likely to look at markers on the map between their areas of residence and
their places of work.
Two such clusters are depicted below, this time including all links within the sub-group (i.e. when
an individual opened a marker or markers created by the other at least once). The first, Figure
8-10, shows the links amongst four people living to the north west of the places of employment,
demonstrating that all links but one (Mike-Ben) were reciprocal. Three of the four people were
involved in the second „most read‟ interaction on the website (Staple Hill to Frenchay Campus),
which is analysed in the next section. Once again, the relative size of the circles provides an
indication of whose markers were opened the most within the sub-group (although this may give
a slightly misleading impression as any response within a thread was attributed by the system to
the person who started it).
Figure 8-10: Who opened whose markers – North West cluster
Figure 8-11 shows the more complex network of connections between a group cycling from the
south west of the employment area.
Findings, Phase 2 – Descriptive analysis
Figure 8-11: Who opened whose posts: South West cluster
This shows a high degree of interconnections and reciprocity within the group, and once again
shows the „popularity‟ of Marion‟s posts (or the threads which she began). Four people in this
cluster took part in the most read thread on the website, which is analysed below.
Detailed analysis of two interactions
This section now returns to sub-question 4.1:
4.1: For what types of informal travel (cycling) information is the case-study system used?
Of the 29 threads, the eleven most popular (i.e. the markers which were most frequently
opened) all started as routes with comments, which then attracted further suggestions, questions
and feedback from other participants. The two most read markers were entitled Horfield to UWE
Frenchay Campus (opened 37 times), and Staple Hill to Frenchay Campus (opened 33 times).
Each of these will now be briefly discussed in order to give a flavour of the website interactions.
Findings, Phase 2 – Descriptive analysis
Staple Hill to Frenchay Campus message thread
In the first example, shown in Box 8.1, three people (Julie, Mike and Ben) discussed a route in a
short thread, and these three opened the marker considerably more than anyone else.
Figure 8-13 shows a screenshot of the opening message. The number of times the thread
was opened by each participant is shown in Figure 8-12. At interview, Julie, Mike and Ben all
remarked that had found the interactions between them both informative and enjoyable.
Findings, Phase 2 – Descriptive analysis
Figure 8-12: Viewings of the "Staple Hill to Frenchay Campus" thread
Viewings of the "Staple Hill to Frenchay Campus" thread
Alice, 1
John, 1
Elaine, 1
Doug, 1
Julie, 10
Inner circles show
participants in the
Ben, 9
Staple Hill
to Frenchay
Mike, 10
Outer circle shows
which participants
opened the marker
and how many times
However, this clustering was not observed consistently; the clarity of the data in this respect
might be obscured by the fact that some participants tended to look at most markers as they
were added throughout the project regardless of geographical relevance, others looked at a
variety of markers at the beginning but then returned less frequently to the site as the project
continued, whilst the possibility of accidentally clicking on a marker cannot be ruled out (a caveat
which in fact applies to all areas of analysis in this section).
Findings, Phase 2 – Descriptive analysis
Figure 8-13: Screenshot: Staple Hill to Frenchay Campus
Findings, Phase 2 – Descriptive analysis
Box 8.1
Staple Hill to Frenchay Campus
This is the quickest and most scenic route I have found so
far. Approx 15 to 20 mins. It does have some hills - 1 short
v steep going up Pearces Hill, 1 longer not as steep going
past Holiday Inn up a lovely country lane.
Staple Hill to UWE: I start at Mail House pub, go down
Pendennis and Croomes Hill, but turn right at double mini.
This takes me along Overndale (short hill) turn left on to
Cleve Wood Rd. There is a hill up to Beckspool but reckon
less than Pearces Hill. If you go round to the right on
Beckspool and turn into Belfields Lane, 3 cut throughs (not
for racing bikes) take you to the start of Filton Rd. Potential
awkward point is crossing Bristol rd, but this route seems
hilly than doing Pearces Hill (or the return journey which is
Tried Ben’s route via Pearce's Hill yesterday and today and
was suitably impressed. Definitely more hilly than the
ringroad route but that's what gears are for (!) and a good
5 mins quicker (which can make a big difference to me in
the morning!) Will give Mike's route a try too - I had no
idea there were so many alternatives!
To: Julie
Hey Julie,
Did you try the Beckspool route? What did you think? I
agree it could do with the plant life being cut back.
Findings, Phase 2 – Descriptive analysis
In this interaction there is a mix of both factual route details (particularly from Mike) and
subjective commentary, conducted in a „chatty‟ style. Ben adds to the route plan he has marked
on the map by alluding to its aesthetic aspects: “the quickest and most scenic route I have found
so far” (….) up a lovely country lane”. Mike‟s response offers informal advice on “3 cut throughs
(not for racing bikes)”, again providing the sort of personal, subjective information previously
found to be typical of word-of-mouth exchanges. Concerns with hills and steepness were
commented upon more than anything else on the website, and are a strong feature of this
interaction. Mike later remarked in interview that he got the impression that Ben was
considerably fitter than he was, as he seemed happy to cycle up hills which he himself avoided.
He also thought that it might have been potentially off-putting for a novice cyclist if he or she had
followed Ben‟s advice. Julie was not put off, however, and her contribution to the interaction
shows that she tried the route straight away and was “suitably impressed” (which Ben found
gratifying). Mike returned to the thread three weeks later to ask Julie whether she had tried his
route, referring to a comment she had made in another thread about the dangers of scratchy,
overhanging plants to the passing cyclist. Because of the time-lag, Julie may not have noticed
this question. As in other exchanges between these three participants, it is notable that they
referred to one another by name, creating a friendly tone.
Horfield to Frenchay Campus message thread
The content of this second example is shown in Box 8.2, with a screenshot of the opening
message provided as Figure 8-15. The four participants in the thread (all located in a
geographical cluster) also opened the marker more than anyone else, but in this case many
other people also read it (Figure 8-14). This may partly reflect the fact that more participants
lived in this part of the city so it was likely to have been of wider interest. However, not everyone
opening this marker lived in this part of the city.
As previously noted, the popularity of the Horfield to Frenchay Campus thread may be partially
attributed to the fact that the highest concentration of participants lived in this part of the city; the
thread also involved descriptions of the two main cycle routes to the employment area from
residential areas to the southwest.
Findings, Phase 2 – Descriptive analysis
Figure 8-14: Viewings of the "Horfield to Frenchay Campus" thread
Viewings of the "Horfield to Frenchay Campus" thread
Adam, 1
Mike, 1
Julie, 1
Sally, 5
Elaine, 1
Doug, 1
Sue, 1
Phil, 1
Jess, 1
Marion, 5
Horfield to
Rick, 2
John, 2
Inner circles
participants in
the thread
Kate, 4
Henry, 3
Alice, 4
Rachel, 4
Outer circles
show which
opened the
marker and how
The Horfield to Frenchay Campus message thread exemplifies the combination of
instrumental/geographical (formal) and social/experiential (informal) information identified
previously in this chapter. The first contributor marked her route on the map and used the
comments box to add a few words about her evaluation of the route: “hard work uphill and scary
coming down because the road surface is so horribly bumpy”. The impression given is of
someone who may regard him- or herself as being a novice, not especially fit, slightly nervous,
cyclist, and who is willing to admit this. This kind of tone may have been appealing to the
Cycology participants overall, as many described themselves in the interviews as being not
particularly confident or experienced at cycling. Interestingly, the people (both male and female)
who highlighted Marion‟s contributions as ones which they especially liked, assumed her to be
male (her website name did not identify her gender). It would seem, therefore, that assumptions
were not being made associating lack of confidence with the female cyclist! The interviews
suggested that when assumptions were made about the gender of other participants (which did
not occur often, as most people‟s website names clearly denoted gender), this was done not so
Findings, Phase 2 – Descriptive analysis
Figure 8-15: Screenshot: Horfield to Frenchay Campus route
Findings, Phase 2 – Descriptive analysis
Box 8.2
Horfield to Frenchay Campus message thread
This is my usual route to work, although sometimes I go
via the MoD site for a change. The big hill in Lockleaze
(Constable Rd) is the least pleasant bit - hard work uphill
and scary coming down because the road surface is so
horribly bumpy. At the UWE end I use the Lockleaze cycle
path (this may not be clear from my route as I couldn't see
where it is on the map!)
Try going straight over rather than right up constable rd -
gentler hill, less busy (see montpelier to uwe route)
how do you go via the MOD site?
To get to UWE via the MOD, go straight on at Locklease Rd
(right would take you to constable rd). Keep going straight
along wordsworth rd, until the next stop sign- turn right
onto Bonnington Walk, over the railway bridge, then turn
left into the football grounds. Follow this path, past the
double metal gates (if they're shut you can go around
them on either side), this will take you out onto the cycle
and pedestrian path. Go straight on, the MOD on your left,
and a housing development on your right.
To get to UWE follow the cycle path past the yellow "gate",
with the nursery on the right. Take the 2nd exit at the
round about (to the right up the hill), follow the cycle path
which should take you to the H.P. entrance next to the
UWE sports ground.
I usually cycle the Lockleaze cyclepath way as it's quicker
(I think anyway!) But if you're cycling late at night in the
dark, the MOD way usually has more poeple about (playing
rugby etc..) and some lights so for me feels a little safer
Findings, Phase 2 – Descriptive analysis
much on the basis of the content of the individuals‟ website contributions, as in a general sense
that more men than women cycle generally.
In response to the opening comment, Sally helpfully offers an alternative to the “uphill, scary”
section and asks about the alternative route mentioned by Marion. At this point Alice writes a
detailed, factual description of the other main route to the university campus from this direction.
This was the only case throughout the project of someone describing a route in words rather
than plotting it on the map. Alice may have done this because she had already plotted part of
this route on the map a few days earlier. She included an informal comment about what to do if a
gate along the route was locked, and this was, in fact, a topic of discussion in another thread
also started by Marion (the “Gates” thread turned out to be the most read „non-route‟ marker).
Finally in this thread, Rachel added her comparison of the two routes, introducing a common
topic on the site: lighting along the routes and her perception of relative safety. Overall,
therefore, this interaction provides a clear example of the way in which word-of-mouth may
comprise both factual (formal) and subjective (informal) information, with the latter adding an
extra layer of personal detail with which the routes might be evaluated.
Chapter Summary
This chapter has reported the findings of the descriptive part of the analysis which sought
specifically to answer research questions 4.1 and 4.2, as well as providing a context for
answering subsequent questions relating to: influence on attitudes, intentions and behaviour;
social-psychological processes; and exploring wider implications in the field of advanced
traveller information. As Phase 2 remained partly inductive, „space‟ was also left for
unanticipated findings to emerge.
Question 4.1 had asked what types of travel (cycling) information were provided and requested
in the case-study system, with reference to the categories identified in Phase 1 (general options
and ideas, factual details, and subjective advice). On the Cycology website, specific factual
detail was usually combined with subjective advice, but general options and ideas about
transport were absent because information was specific to a „niche group‟ of users (cyclists) who
were seeking and offering specific rather than general travel information. Many posts combined
formal information such as factual route descriptions and lines on the map with subjective
evaluations of the experience of cycling along them (informal information), supporting the Phase
1 findings concerning people‟s preference for integrated formal and informal information. The
main themes appearing on the site were wayfaring (geographical route descriptions), safety,
cycling infrastructure, sharing space with other road/path users, and subjective (affective)
descriptions of the cycling experience.
Findings, Phase 2 – Descriptive analysis
Trends in content over time were shown by grouping posts into broad categories according to
subject matter (route or non-route related) and function („pro-active‟ statement or question;
„reactive‟ response). It was shown that route information constituted the majority of posts, and
most of these were posted during the first week. Questions also peaked in the first week, but
responses to questions only peaked in the fifth week. Non-route related comments and
responses to them both peaked in the fourth week. Overall, the offering and requesting of nonroute information, as well as general comments and responses to them, were relatively steady
over the six weeks, although small in number and tailing off at the end. Although this kept the
website interaction „ticking over‟, the dwindling numbers suggests that „injections‟ of new
participants, information and opinions may have been required to maintain the level of postings
at one or more contributions on most days if a project such as this were to continue for a longer
period. This frequency of posting allowed a level of interaction which was regarded as „sufficient‟
by most participants (to be returned to in Chapter 11).
Question 4.2 concerned word-of-mouth processes and patterns of interaction. It was shown that
geographical clusters of people travelling from the same areas tended to interact with one
another on the website (unsurprising as the majority of posts related to routes), but there was no
such pattern among the interactions dealing with non-route related cycling matters. In terms of
the markers opened, there was also a slight tendency for people to focus on reading posts
created by people who lived in the same area, but many also read posts by other people from
other areas as well. It was also demonstrated that most people wrote and read similar
proportions of posts, with few obvious „information-givers‟ or „information-receivers‟. The role of
different individuals has helped to inform the development of a participant typology, to which we
return in Chapter 11. The next chapter will focus on the influence of Cycology on participants‟
attitudes, intentions and behaviours, whilst Chapter 10 will move on to the role of socialpsychological factors.
Findings, Phase 2 – Behavioural influences
Chapter 9 : Findings, Phase 2 - Influence on attitudes,
intentions and behaviour
This part of the analysis aims to answer the research question 4.3, based primarily on a thematic
analysis of interviews and questionnaires, and supported by observation of the website.
4.3. What influence (if any) has word-of-mouth information exerted on participants‟ selfreported attitudes, intentions and travel behaviour within the case-study?
This section presents four areas where influence occurred: the trying out of new routes seen on
the website; becoming more active on cycling issues (complaining/campaigning); reinforcing procycling attitudes, intentions and behaviours; and changing the experience of the commute. A
context is provided by the Phase 1 findings in this area, which showed that information obtained
through word-of-mouth was reported to have influenced participants‟ trip details but not more
„strategic‟ decisions such as modal choice (although some believed they had influenced the
modal choice of others). It was found that normative information obtained through general
conversations was thought to have had no direct effect on trip choices, but may have influenced
general attitudes. Overall, participants had described a process whereby informal and word-ofmouth information had been combined with formal types and sources, and used as part of a
rational (individual) decision process consistent with information processing theory (Bettman,
Trying out new cycle routes
The most significant effect of the Cycology website in terms of altering behaviour - and the one
which was the easiest to identify – related to route choice amongst existing commuter cyclists
(who, as has already been explained, accounted for all but one of the participants). The
interactions on the website showed that four new routes, recommended by others, had been
used, and two people also reported their intentions to try out a route they had seen on the site.
One such example on the website said:
Tried out the suggested North View/upper Cranbrook Road route last night,
and it was way better than my normal route apart from the horrendous right
turn onto upper Cranbrook Road, but I’m sure I’ll get used to that! (Rick) 9
Verdana font continues to be used to denote verbatim comments from the website.
Findings, Phase 2 – Behavioural influences
As in previous and subsequent chapters, is used to (italicised, Arial font). However, the
interviews and post-project questionnaires revealed that there had actually been considerably
more trying out of routes than the website postings suggest, possibly because some routes were
used after the end of the project, or simply because people did not „get around‟ to giving
feedback on the site. Figure 9-1 below shows that, in total, more than half (13) of the participants
reported having tried a new cycle route suggested by another participant, two had re-tried an
„old‟ route (i.e. a route they knew but had not used for a while) and a further three said they
intended to try out a different route when circumstances permitted. For one person these
circumstances included trying to remember to veer from his habitual route:
“I was going to try altering my route and go the way someone had indicated on the map,
but I found I was on autopilot on my normal route and kept forgetting to try it, old age I
suppose!” (Jim)
Figure 9-1: Effect of shared information on participants' route choice
Effect of shared information on participants'
route choice
Number of participants
Because the choice of cycle routes from one location to another within a few kilometers is
limited, what was described as a “new route” was often in fact a variation to part of an existing
Findings, Phase 2 – Behavioural influences
route. Two of the four people who did not consider changing their route reported the reason as
being lack of any relevant suggestions or better options, for example:
“Well you know, if somebody had…. if there'd been an interesting route that would have
worked for me, then maybe I would have tried a different route. But there wasn't really
anything that served my purposes, I don't think”. (Doug)
“(There‟s) only one other route pertaining to Bradley Stoke and I think it‟s suicidal”.(Elaine)
The longer-term behavioural effects of gaining knowledge of, and trying out, new routes varied
according to participants‟ evaluation of the routes once they had experienced them. Sometimes
it merely led to a confirmation that someone‟s habitual route was „the best‟ for them:
“I did re-try an old route that I used to use to get to work, prompted by reading some of the
comments about that route on the site. However, I soon remembered why I stopped using
it previously (too many obstacles), and went back to my usual route!” (Marion)
For another person, speed was the most important factor, and having tried an alternative route
from the website, he was confident that his usual route choice was indeed the fastest:
“And I'd always wondered. "am I missing something, is there a route that‟s quicker that
I've just not cottoned on to?" It was good to know that that was not necessarily the case,
and that the routes I was taking were the fastest routes”. (Ben)
Others reported that information from the website had helped them to widen their repertoire of
routes, which meant that they had moved from always taking the same route to varying their
route on some days, or they simply kept a „bank‟ of alternative routes in mind:
“And I definitely tried them and thought "yeah, this is good". And I've gone from one
route, which I was kind of doing day in day out, which is a bit boring, to varying it now,
and some days I just sort of, on the fly, think "oh, I'll go that way instead". And I do
different route”.(Kate)
For some however, the website had provided a new route which turned into a participant‟s
regular route choice, for example:
“‟Though it was useful for me. The two people that I did communicate with, who came
from the South Gloucestershire side, I did try both their alternative routes. There‟s one
of them I'm sticking to now, as it's much better than the one I was doing before”.(Julie)
Findings, Phase 2 – Behavioural influences
The data therefore provided evidence – as far as is possible for reported behaviour - that the
sharing of route information on the Cycology website did have an impact on the route choice
behaviour of the majority of participants. This corresponds with the Phase 1 findings about the
influence of word-of-mouth on particular details of a trip (in this case, the route). However, the
findings in Phase 2 suggested a higher degree of social influence (albeit in the narrow field of
cycle route choice) than the more general accounts provided by participants‟ in Phase 1 had
indicated. This might be explained by the high levels of trust within the small experimental group
(considered in Section 10.1) and by the fact that trying a new cycle route is perceived as „low
risk‟. It may also imply that social influence is more prevalent than many people may wish to
admit when reflecting on their general travel behaviour, and may be particularly prevalent within
a salient reference group: in this case a group of fellow cyclists and co-workers. This explanation
will be discussed in detail with reference to social-psychological theory in Chapter 10.
Reinforcing pro-cycling attitudes, intentions and behaviours
Although there was no evidence from the interviews that participation in the project actually
changed people‟s attitudes towards cycling or travel in general, it seemed that it did, for some,
serve to reinforce what one person described as her “cycling is good ethos”. A participant who
had recently started cycling said that her involvement in Cycology had encouraged her to
“Being part of this project did improve my enthusiasm for cycling, and as it went on just as I
have started cycling to work, it encouraged me to keep at it”. (Alice)
Attitude theory suggests that attitudes can strengthen when supported by direct and indirect
experience - in this case, the experience of cycling. The associations which link the attitude
object to relevant prior experience create an intra-attitudinal structure - the internal structure of
the attitude (Eagly and Kulesa, 1997).
Alice and another participant both commented that this sense of encouragement partly
emanated from realising that many people cycle further than they do:
“I suppose I was already thinking that cycling was good, and the way to get around.
And I suppose it might have reinforced it a bit. You know, that there are other people
out there doing all sorts of different routes, and from further away, as well. So I was
thinking, you know, my half-hour seems like plenty, then you see lots of people who are
doing further routes, and you think "that's great, good for them!". You know, it kind of
reinforces that my half hour isn't that bad, and even in winter, it'll be fine - that sort of
thing”. (Sally)
Findings, Phase 2 – Behavioural influences
This can be interpreted as a change, for this participant, in the descriptive social norm relating to
the „normal‟ distance which other people cycle (Cialdini et al., 1990).
It was perhaps of note that both these women had started cycling to work within the previous six
months. A participant who had started cycling to work within the previous year also thought that
the project had served a motivational purpose, particularly because she had discovered a better
route. So even if it had not changed her overall attitude to cycling, it might have made her feel
more positive about her regular commute.
“I'm not sure that it changed my views like that particularly. I suppose I was more
encouraged to cycle, and I think I probably cycle more now that I've found myself a
route. Maybe, yes. Maybe it's improved my opportunities for cycling.” (Helen)
Another participant who had started to cycle only within the previous year, who did so only
infrequently, and described herself as a nervous cyclist, remarked that reading about other
people‟s experience of a particular route, which involved a “frightening” road crossing, had given
her the confidence to try it.
“I think it does make you think about things, if you've got a conversation going on, that
maybe I would like to be a part of that. This last week I've been so frustrated. Sat in the
car. And I'm actually thinking "for God's sake, it's only a bit of dark. Get a light. Sort
yourself out ". (Esther)
This participant particularly enjoyed the social aspects of the site, remarking that it had helped to
increase her resolve to cycle more often and not to be put off by the darker mornings, even if this
intention had not yet materialised into action (exemplifying the widely identified „attitudebehaviour gap‟ in social psychology; e.g. Ajzen and Fishbein, 1977; Anable et al., 2006). This
might also be interpreted with reference to the Theory of Planned Behaviour (Ajzen, 1991): the
participant‟s perceived behavioural control may have changed because cycling on dark
mornings is now thought possible if a plan is instigated to obtain some bicycle lights.
The one person in the project who was not yet cycling to work thought, hesitantly, that it had
provided her with a little encouragement to start, hence reinforcing an intention:
“It was nice to see other people cycling from near me, and even further away from me. I
suppose it might have sort of reinforced my desire to cycle. To work. At some point. A
little bit”.(Helen)
She thought that she would have derived greater encouragement from the website if there had
been more participants in the same position as herself:
Findings, Phase 2 – Behavioural influences
“Yeah, just more thinking "there‟s other people in the same boat".
Or "look we‟re
both…". Maybe someone could have been like "well, I didn't ride to UWE yesterday, but
I rode from my house to this point on the map, and that's good, and maybe I'll work
myself up to going to UWE…". If there'd been more conversation like that, it could have
been like "oh yeah, I'm going to try that this week!" or whatever”.(Helen)
This suggests that positive attitudes towards, and intentions to, cycle are more likely to be
reinforced when there is some sort of identification with participants in the interaction, a
proposition which will be explored further in Section 10.2: „community‟, in the context of self
categorisation theory (Turner et al., 1987).
It would seem, therefore, that the participants who were most susceptible to encouragement, or
most likely to have positive attitudes to cycling reinforced through the online interactions, were
those who were relatively new to cycling to work. This may be because perceived barriers to
cycling (such as distance, effort, wet weather) may still be salient in the mind of someone who
has recently switched from another commuting mode, but these barriers may reduce in
significance as he or she becomes more accustomed to it. Participants who had been cycling to
work for many years remarked in the interviews that „social support‟ for cycling was not
something they particularly required, because cycling was simply part of their routine; beliefs and
attitudes were strongly in favour of cycling based on considerable experience, and were thus
more stable and less likely to be influenced by others. However, many people in this group
thought that they would have found the website helpful (in terms of both practical information
and social support) when they had just started cycling, and might even provide an
encouragement to those who were not yet cycling “because I think there‟s a lot of people out
there that could cycle, that don‟t cycle, partly because they might not know what route to take.
They might not know that there‟s a culture of people that do cycle as well”.(Ben)
This moves the focus from the actual effects of the project to participants‟ thoughts about how
the project might – hypothetically - have affected their attitudes, intentions or behaviours under
different circumstances, or how they thought it might affect those of other people.
“I think one of the common barriers to people getting on the bike is, they see the busy
road, like they say "I don't want to cycle down the Gloucester Road." So something like
this can show them an alternative. Because if you just look on the map, you can't see
that there is a way through behind the MoD for instance (….) Whereas if you look here,
you see it there is another route, and you think "other people do it". .(Phil)
Findings, Phase 2 – Behavioural influences
“…if I was new to cycling, and I was looking at something like this, and I was looking for
support, I might look at that and think "okay - other people go through these situations,
and they feel good about it and stuff”. So I won't be quite so put off by the fact that I got
drenched last night ". (John)
This sense of encouragement from the knowledge that „other people do it‟ exemplifies the type
of background, normative influence on attitudes and intentions - effected through social
interaction - which emerged from the Phase 1 interviews.
Taking action on cycling issues
Some participants felt that their involvement in the project had spurred them into taking some
sort of action as individuals, if not as a group. This was reported to have happened for two
reasons: firstly, if you notice a problem along a cycle route and then tell other cyclists about it, it
causes you to think about it more seriously; and secondly, if others also remark that they are
experiencing the same problem, you feel that you have stronger grounds for making a complaint.
For example, one participant explained that she had complained to the Council about a poor
road surface after she had commented upon it on the site, whilst another, who saw few other
cyclists at the early hour he travelled, was motivated to take up the issue of a locked gate after
being obstructed by it himself and then seeing it discussed on the website:
“So, that was quite strange, to actually stop on the route and realise how many other
people were using it at the time. But it was also a bit scary that they closed the gate…..I
spoke to the guy from South Gloucestershire, who is planning the new cycle routes, and
mentioned to him that they'd closed it and said "surely that's the key route for lots of
people." (Phil)
At one stage in the project a question was posted on the website about the lack of lighting on a
well-used cycle path:
My main issue is the lack of lighting on the cycle path - makes for a
dangerous ride in the winter. Any ideas whose responsibility this would be?
However, this was not answered, leaving the contributor feeling (as she said in the interview)
that “perhaps I‟m the only person who thinks the path is dark” - although there were in fact other
mentions of this concern on the website. One person asked what others‟ approach to using the
“dark and scary” cycle path was, and another said in interview that she had intended to reply but
had not got round to it.
Findings, Phase 2 – Behavioural influences
Interestingly, and in contrast to the general view, one participant remarked that the presence of
contributors on the website who seemed concerned to „get things done‟ actually made her less
likely to take any action herself. This could be interpreted as a free-riding effect. Commenting on
a situation where she had noticed that a particular set of traffic lights - one which is usually
sequenced to give cyclists a chance to cross a busy junction on their own – were not functioning
properly, she said:
“And I just mentally thought, partly, again, seeing other people on there who were much
more active about talking about things and doing things than I am. I thought, "there'll be
another cyclist, who will be on to that really quickly. Who‟ll ring somebody up and say
the cyclists‟ lights are out - sort it out". I mean, that's a bit lazy on my part, it just gave
me the sense that - a lot of people who cycle are quite proactive. So if I hadn't done that,
I may have just thought "gosh, I hope someone does that". But I had more of a feeling
of "oh yeah, someone will do that". (Kate)
As well as actual effects, the issue of the potential campaigning effects of the website also arose
in some interviews. Many people mentioned spontaneously that a website like this could be a
good way of communicating with local councils about cycling issues, either by allowing transport
planners and cycling officers to access the site, or simply by encouraging people to act
individually or band together to contact the council about shared concerns. It was felt that the
latter would probably have happened naturally, if the project had continued for longer (in fact, the
cycling layer of the main bristolstreets website had hosted a cycling survey in which users were
invited to mark issues of concern on the map for the attention of the City Council).
It was also suggested that a website of this nature could give Councils an insight into what
„normal people who cycle‟ are thinking, as opposed to the sometimes more radical views
expressed by the cycle campaign organisations. This is interesting in the light of Spinney‟s
findings about the „professionalization‟ of cycling stakeholders and corresponding
disenfranchisement of the „non expert cyclist‟ (Spinney, 2009a). One person articulated the
potential of the website in this area as follows:
“I think after the routes went up, people started making comments about the different
traffic habits, the state of the road and things like that. That was really useful, and I think
that might be something that perhaps Bristol City Council or South Gloucestershire
Council might want to take a view on. Be a part of, perhaps, and be able to view some
of those things, because it's a bit of an insight into what works for cyclists. Traffic
planners should have access to this (laughs), definitely” . (Julie)
Findings, Phase 2 – Behavioural influences
Changing the way people experience their commute
Half of the participants reported more unexpected, psychological effects from the project, saying
that it had subtly changed the way they experienced their everyday commute. Here, the focus is
not on tangible effects, such as change of route, but on the „lived experience‟ of cycling, or what
Spinney (2009b) describes as the “fleeting and ephemeral meanings that arise through cycling
as an embodied and sensory practice” (p.829).
All of these people cycled to work either every day or frequently. Some said that they had
thought more carefully about their everyday cycle ride as they looked out for features
commented upon by others who rode the same route (such as the black cat) and also paid more
attention in looking for interesting features and stories which they could describe to others. Two
people in this group took photographs along their routes (but unfortunately were unable to
upload them), and one person experimented with some new routes, which had not been
provided by other participants, just so that she would be able to tell people about them.
A number of people remarked that their participation in the project made them feel less isolated
during their cycle commutes; even if they actually saw very few other cyclists along their way, it
gave them a sense that lots of other people were going the same way at different times of day.
“It was good to kind of…. I suppose as a cyclist, you feel fairly isolated anyway, you
don't realise that other people are cycling in. So it's useful to have someone else who
cycles to chat to, and it's nice to know that you're not the only one that‟s cycling in. So it
was really useful. It was good.” (Ben)
“Yeah, it just opened my mind up to think "well there's other people doing the same
routes, or similar routes." And they go through similar things I go through (….) It made
me think "but we‟re all having quite common experiences here". And it just made me
feel a bit more like I wasn‟t alone doing the things I'm doing on my commute. So it was
quite nice” .(Kate)
One person described this as „personalising‟ the cycle journey:
“The fact that it does feel very lonesome to set out on your first cycle ride to work. You
don't know who else is doing it, you don't know if anyone else is going the same way as
you. If you've got a few people on there and you've read that and gone " these people
do this every day, they're setting out earlier than me." (.…) I think in that way, what it
really does is it personalises cycling, and it makes you feel very confident in the fact that
other people are doing it” .(Adam)
Findings, Phase 2 – Behavioural influences
In a similar vein, this participant remarked that the project had given him “quite a nice feeling
about cycling in that area, and it does make you feel quite like you own the road, as if it's your
area, in some ways, because people are going through it and it's yours.” He described the
experience as one of „wayfaring‟, and of a sense of his cycle trip becoming „humanised‟ by the
awareness that others were using it:
“It showed me the diversity of routes that are available and also I quite like the „wayfaring‟
feel that the site engendered. It humanised the cycling route as you were able to see other
people‟s stories posted in places that I go through every day and you could think about that
as you were cycling through and imagine all the other people with different experiences and
stories of the same place that use it every day”.(Adam)
Chapter Summary
This chapter has sought to answer research question 4.3 concerning the influence of word-ofmouth information on participants‟ attitudes, intentions and behaviours. The effects arose from
both formal and informal types information, the key factor being that it was communicated
through word-of-mouth amongst people within the group. The findings ranged from the direct,
tangible effect on behaviour in the form of people using routes they had seen on the website, to
more subtle psychological effects on the way in which people experienced their commute by
bicycle. It was thought to have strengthened pro-cycling attitudes and intentions through direct
and indirect experience (intra-attitudinal structure). The effect on attitudes corresponds with the
Phase 1 findings that social interaction can exert a „background influence‟ on attitudes to travel.
When it came to making active travel choices, in Phase 1 it was clear that participants believed
word-of-mouth information to be just one influencing factor within a complex process of
individual, rational decision-making (depicted in the lower part of Figure 6-1, p.95). By creating
an environment where social influence could be directly observed, Cycology demonstrated that
people may be more susceptible to influence from others than they might generally realise, or
admit to. This effect may have been accentuated by the context of the small group (reference
group) within which the Phase 2 participants were placed. However, it was also clear that in
noting new routes and trying them out, people were engaged in a rational, individual decision
process, and were not simply following others without deliberation. Therefore, the new
knowledge generated by this research about word-of-mouth influence seeks to complement, but
not replace existing models of the role of information within the decision process. A process
model developed from the current findings is presented in 11.5. First however, the following
chapter will discuss in detail some of the key social-psychological mechanisms which were
found in this research to explain why participants were influenced in the ways they were.
Findings, Phase 2 – Role of social-psychological factors
Chapter 10 : Findings, Phase 2 - The role of socialpsychological factors
The objective of this part of the analysis was to explore the ways in which social-psychological
mechanisms functioned in this online environment, based on a thematic analysis of interviews
and questionnaire, supported by interpretation of observed behaviour on the Cycology website.
Would information from others be considered trustworthy, and if so, why? Would social factors
such as social proximity (membership of an „in-group‟) influence levels of trust, or would trust
appear to be based primarily on instrumental factors such as the informant‟s direct experience of
the route? And what would motivate people to offer information to others? Drawing on socialpsychological theory from the literature review as well as the Phase 1 findings, this part of the
analysis sought to answer the research question 4.4:
4.4 How do social-psychological mechanisms appear to function in this online context?
A number of more specific areas of interest arising from the Phase 1 exploratory research
were identified for exploration:
4.4.1 Obtaining word-of-mouth information from others:
In Phase 1, trust was found to be a key factor affecting informal information-use. Whose
information did people trust in this online environment, and why?
Relational trust:
 Familiarity (e.g. knowing someone „off-line‟ as well, or coming to know them
through regular online interactions)
 Similarity (e.g. sharing socio-demographic characteristics, similar attitudes to
transport, personal norms, levels of fitness, or shared interests)
OR social proximity through being within the same community (e.g. within an
organisation or locality)
Calculus trust:
How would this interact with an „instrumental-reasoned‟ trust in the other person‟s
What might make information seem less trustworthy?
Findings, Phase 2 – Role of social-psychological factors
4.4.2 Giving information to others:
Why did people contribute information to the case-study system? The following issues from
Phase 1 would be explored:
Pro-social behaviour/ reciprocal altruism/social exchange
Are people motivated by their membership of an in-group (e.g. within an organisation or
What might prevent people from posting information?
In order to maintain the flexible design and open nature of the research, these questions were
defined not with the intention of providing a fixed structure for the findings, but rather as a guide
to help identify the potential areas of interest which might emerge. Therefore, answers to these
questions are not reported sequentially, but rather, the findings will be presented as a number of
interlocking themes which emerged from the coding and analysis, with reference to specific
questions where appropriate. The themes are displayed in Figure 10-1. They are: trust; group
identification; small group effects; social judgement and pro-social behaviour. These themes will
be addressed in Sections 10.1 to 10.5 below.
Findings, Phase 2 – Role of social-psychological factors
Figure 10-1: Interlocking social-psychological themes
Interlocking Social Psychological Themes
Section 10.1
Section 10.2
Section 10.3
Small group
Section 10.5
Section 10.4
10.1 Trust in the information posted
This part of the analysis principally addresses research question 4.4.1, which relates to trust in
the information received from others. In Phase 1, trust was found to be a key factor affecting the
use of information provided by others. Trust was found to be rooted mainly in the real experience
of the information-provider („local knowledge‟), but was also found to be affected by socialpsychological factors such as social similarity with the information–provider. Most of the Phase 1
participant accounts concerned face-to-face interactions with individuals they knew. The
question to be answered now was whose information would people trust in this online
environment where the people did not know each other, and why?
Questionnaires and interviews revealed that all respondents believed the information posted on
the site to be reliable and trustworthy. This was also suggested by the generally friendly and
harmonious tone of the interactions observed on the website, and the positive responses when
people had followed a suggestion or expressed an intention to do so. Most of the interview and
questionnaire responses indicated a calculus-based reasoning for this trust relating to the
intrinsic quality of the information: because participants had real experience of the routes;
Findings, Phase 2 – Role of social-psychological factors
because the information was up-to-date and inaccuracies could quickly be corrected by others;
and finally, because a level of detail could be provided which was absent from other (formal)
information sources. It will be recalled from the literature review that calculus-based trust is
based on rational choice and is typical of short-term interactions where the truster must calculate
whether the trustee intends to behave in a way which is beneficial to the truster. Trust is derived
in part from credible information regarding the intentions or competence of the other person
(Rousseau et al., 1998). Many participants were also able to compare some of the posted
information with their own experience, or were prepared to test the reliability of information by
simply trying out a suggested route. This supports the Phase 1 findings that calculus-based trust
is the primary trust mechanism when people assess the credibility of word-of-mouth travel
However, some participant responses implied a more „social‟ dimension to this trust in the routes
and comments in Cycology because: “actual people „with faces‟ had posted them”. Sometimes
this involved a judgement being made about attributes (especially fitness level) of an individual
information-provider, but more usually trust was based on assumptions about the good
intentions of the group as a whole. This corresponds with the concept of relational trust, which is
more likely to derive from interactions over time between the different parties, where trust is
based on information available to the truster from within the relationship itself, within which
emotion may play a part (Rousseau et al.,1998). These different contributory factors are
summarised in Figure 10-2, with calculus-related factors on the left of the diagram, and relational
factors on the right. Each of these factors will be discussed below.
A theoretical dimension to the understanding of trust within the small group of Cycology
participants is provided by theories of social influence. Deutsch and Gerard‟s dual process
theory (1955) suggests that members of a group are more likely to take the judgments of other
group members as trustworthy evidence of reality (compared with non-group members), and,
hence, are more susceptible to informational social influence. In the literature review (Section
2.3) it was explained that informational influence is based on the acceptance of information
obtained from others as evidence about reality, whereas normative influence is based on the
need to conform with the positive expectations of others, particularly in a group environment.
Turner (1991) later identified a form of social influence called referent informational influence,
whereby people adjust their identity, attitudes and behaviour to correspond with the collectively
defined attributes of their social groups (Wetherell, 1996). Referent informational influence
integrates the twin concepts of normative and informational influence, asserting that the basic
influence process is one where the normative position of people categorised as similar to self
tends to be subjectively accepted as valid (Turner, 1991). Thus, it is not the informational
Findings, Phase 2 – Role of social-psychological factors
content per se of others‟ opinions and actions which matters, but the extent to which it is
validated by ingroup consensus (Turner et al. 1987). In the following sections consideration is
given to how far these theoretical assumptions were borne out in the Cycology study.
Figure 10-2: Factors contributing to trust in the information posted on Cycology
wants to be
Experience of
‘real people’
with past
with others’
about (good)
intentions of the
quality of
Reputation is
important in a
small group
reliability of
with own
attributes of
fitness level
to mine
cycling style
to mine
Work/live in
same place
as I do
Test by
trying out
10.1.1 Calculus-based trust
Intrinsic quality of information
Factors relating to the intrinsic quality of the information (shown to the top left of Figure 10-2)
were among the most frequently cited reasons for believing it to be reliable. Among these, the
most common and fundamental reason for trusting the information was the view that it was based
on the real-life experience of „real cyclists‟, and that this was lacking from other sources of
(formal) information. This corresponds with the significance attributed to „local knowledge‟ in
Phase 1 of this research.
Findings, Phase 2 – Role of social-psychological factors
“And I certainly felt that it was reliable in that people had tried using them. They'd kind
of done the route, and it was their route that they took to work, and so I kind of trusted it
from that point of view. And there was nothing else really like that, that I'm aware of,
that you can use”.(Ben)
“I appreciated the „real‟, personal comments that people made – about what they
personally enjoyed or found difficult about a route (…..) I was inclined to trust the
information given because I knew it was from real cyclists….”.(Marion)
Regarding accuracy, another participant remarked:
“it becomes a discussion, doesn't it, of things. So in a sense, if somebody put
something that is outrageously incorrect, it's not a bad thing, because it does encourage
other people to refute it. And give you good information”. (Jim)
In this sense, an online group format such as the Cycology website was thought to offer
advantages over one-to-one interactions because inaccurate information could be swiftly
corrected by others, and in cases where opinions differed, the reader might be guided by the
consensus or majority view. Hence, information appeared to be perceived as more reliable if it
represented a group norm. The link made between group consensus and information reliability
suggested a process of accepting agreed norms of opinion within the group. This illustrates
Deutsch and Gerard‟s concept of “normative informational influence” (1955), and here a parallel
can be also drawn with recent literature on the credibility of electronic word-of-mouth in the
context of online consumer recommendations. Drawing on Deutsch and Gerard‟s dual process
theory (1955), Cheung et al. (2009) categorised the consistency and rating of online
recommendations as normative determinants of information credibility, and found these factors
to be influential alongside informational determinants such as argument strength and
confirmation of prior belief. Similarly, one of the reasons why information on the Cycology site
was deemed credible was that it was thought to represent a group consensus (if anything
inaccurate were posted, it was expected to be corrected by others).
Within this category (intrinsic quality of the information), the third factor contributing to perceived
reliability was the level of detail contained within the postings. Most participants believed that the
website provided a degree of detail, based on users‟ experience, which could not be provided by
formal sources such as a printed cycling map, with the added advantage of being kept up-todate. Examples of such details which were thought to enhance the credibility of the information
included: condition of the road; cycle-theft hotspots; broken glass on the road; and pleasant
views. One person noted in her questionnaire:
Findings, Phase 2 – Role of social-psychological factors
“the advantage is that this is up to date and people can tell each other about anything that is
blocked, by road works etc. The gate on the rugby ground that is sometimes shut is another
example. In many ways it is more reliable than the maps because it tells you what is
actually there!” (Elaine)
However, the relative advantages afforded by a standard cycling map in offering far more
comprehensive information was also noted by many.
Comparison with own experience
The second category of calculus-based trust factors depicted in Figure 10-2 is comparison with
own experience, which might take the form of agreeing with another person‟s route directions
based on past experience (a rational comparison), empathising with another person‟s
experience (an emotional comparison), or simply testing out a route suggested by someone else
(“physical reality testing” – a social-psychological construct outlined below). First among these
factors, the findings showed that trust was reinforced if information on the site corresponded with
the reader‟s own experience of cycling the routes which were described. For example,
commenting on posts from other participants which related to his own route, one person said:
“And then, so I…. because I found the ones on my route to be true, I would expect that if
I saw a comment somewhere else, I would take it as read really. I would believe it.”
In this case, it appeared that a heuristic rule was being applied – if what one person in the group
says corresponds with my own view (i.e. it is „true‟), I will expect the contributions of other people
in the group to be true as well. This is an example of a judgemental heuristic, whereby quick and
efficient judgements are made by rules-of-thumb which generally yield valid results (Fiedler and
Bless, 2001). Affective processes might also be involved in this judgement: the recognising of
shared experiences and corresponding feelings of empathy could also increase the trust which
someone might have in another participant‟s viewpoint:
“Of course, you can never really know what people are really like. With an Internet
persona. But you know, if people are cycling to work the same as me, you know that
they're going through the same sort of things. You know if it's cold and wet, or if cars
are screaming by. So you kind of think, because there‟s a shared experience, you
know, you tend to believe them more”.(Phil)
Finally in the „comparison with own experience‟ category, it was also common for people to test
the reliability of the information by simply following a recommendation and assessing whether or
Findings, Phase 2 – Role of social-psychological factors
not they agreed with the advice which had been given. This would then affect the credibility of
other comments from the same person, or, heuristically, other contributors as well. This can be
interpreted as a process of “physical reality testing” (Festinger, 1950). According to Festinger,
people use different methods to test the “subjective validity” of their opinions, attitudes and
beliefs. These methods can be placed on a continuum between physical reality testing on the
one hand, and social reality testing on the other. Where it is possible to test an opinion, attitude
or belief physically (such as riding a cycle route which one thinks may be good), subjective
validity is confirmed or denied through one‟s own direct experience; this was the case when
participants tested out cycle routes which they saw on the website. At the other extreme, where
no physical reality check is possible, the validity of one‟s opinions, attitudes and beliefs can only
be tested by comparing them with those of others (“social reality testing”). The reported role of
Cycology in reinforcing a general „cycling is good ethos‟ (Section 9.2) might have resulted from a
process of social reality testing among some members of the group. Knowing that „similar others‟
shared their positive view of cycling appeared to render some participants‟ positive attitudes
subjectively more valid.
A benefit of an online system is that it allows people to share the results of this reality testing
process with other readers, typically through a „user rating‟ system. One participant suggested
that such a rating system might be useful in this context of online travel information:
“I mean in my opinion, I suppose with any site like that, like with eBay or anything like
that, people gain trust by you experiencing something, and people leaving feedback, so
like rated users: how good is this route, how trustworthy are their comments - rated by
the user community”.(Adam)
This open-minded approach to testing out people‟s advice was facilitated by the view that taking
a new cycle route on someone else‟s recommendation was a low risk activity so it is easy to give
other people‟s suggestions the benefit of the doubt. One person articulated this view as:
“I just thought "that looks useful. Let's give it a go". Because, you know, you try it, and if
you end up at a dead end, you cycle back. It's not disastrous”. (Sally)
This corresponds with Deutsch and Gerard‟s (1955) assertion that greater trustworthiness is
possible where the reliability of other group members‟ judgements can be checked.
10.1.2 Relational trust
The reasons which have been reported so far for participants‟ belief that information on the
website was to be trusted can be all categorised as expressions of calculus-based trust.
Findings, Phase 2 – Role of social-psychological factors
However, it was clear that the high degree of trust could also be attributed to more social factors,
based on assumptions made about other individuals, and, in particular, the group as a whole.
This may be categorised as a form of relational trust, where trust arises from within the
relationship itself (Rousseau et al., 1998). The concept of relational trust has been slightly
amended in its application to Cycology, since most literature in this field concerns interpersonal
trust arising from repeated interactions between two individuals.
Assumptions about the (good) intentions of the group
In Cycology, trust arose more from the relationship between individuals and the group, rather
than between particular individuals. Hence, there was an over-riding belief that everyone on the
project was well-intentioned, and trying to be helpful, leading to a norm of providing reliable
“Knowing that these are real people and that it was a relatively small group, I felt in no way
that I needed to doubt any information. Since this is a voluntary project, I was certain that
the nature of posting information would be correct and useful to the project.” (Laura)
Assumptions were therefore being made about the benevolence of people‟s motivations within
the group (Deutsch and Gerard, 1955). Three constituent factors relating to assumptions about
the group‟s good intentions are depicted to the top right of Figure 10-2. It was assumed that
everyone wanted to be helpful, not only because they had volunteered for the project, but also
because they might learn something useful for themselves; moreover, there was a general
assumption that people who cycle tend to be supportive of one another (described in the next
section on group identification). As one participant noted:
“There's nothing to be gained from putting any misleading information in there.
Everyone is actually trying to help each other really and trying to improve their own
experience of cycling, I suppose. Yeah definitely. It definitely felt very trustworthy, the
messages coming up”.(Rick)
Occasionally it appeared that self-presentation and impression management concerns (Leary,
1995) were also playing a role for some people in ensuring that they provided accurate
information in order to earn a reputation for good posts. It was assumed by these participants
that others in the project would be subject to the same pressure and hence were to be relied
upon. This reflects a type of normative pressure to provide reliable information.
“I think there's a high bit of accountability with this sort of thing (…) I mean, when I was
writing mine I was thinking, I'm posting a good route, I want to give a good description, I
Findings, Phase 2 – Role of social-psychological factors
want people to be able to use my route because it's, you know, because it is a good
route, and I enjoy cycling on it. So I just tended to trust them, I had no reason not to.
Nobody made a false post or anything like that” (Adam)
As expected, the small size and restricted nature of the group - and the consequent sense of
association with other participants which this created - also increased the level of trust which
some people placed in the website information (and in those providing it). An association
between small group size and an „automatic‟ trust in others - despite not knowing the individuals
involved - was made by a number of people. For example:
“I didn't really know anyone in advance. Because it was such a small group of us, in a
sense, (….) well, I automatically trusted them, really, and their advice.” (Rachel)
This was contrasted by some with the lower level of trust they would place in a publicly
accessible website. One person stated: “I personally tend to have much less trust in comments
when I don‟t know who has provided them”, and another commented on the “trusted user-base”
of a small, secure website. Asked about the reliability of the information, one participant
employed at the university mentioned a sense of „reassurance‟ in knowing that most people in the
project worked at the same place:
“To be honest, I think perhaps, the fact that it was kind of the UWE staff that were doing it as
well. I don't know why, but I found that more reassuring, that they were people that worked
at the same place as me. And, yeah, I think when you use that kind of facility to talk to
people, you've got that more one-to-one aspect of it, so ….. that makes it more reliable.”
Another participant thought that, compared with a user-base within the university, where it was
always possible to find out who people were, she would have less trust in the information on an
open website, and would also be more wary of providing information if she did not have an idea
of who the readership might be:
“Whereas when it's just Joe Bloggs, you have no idea who they are or where they‟ve
come from. I think you might be more wary about giving any information to them. Or
really trust their information, before you've used it before”.(Rachel)
However, this point should not be over-emphasised, as most participants thought that they would
also expect information posted on an open website to be reliable, and would generally take it at
face value, albeit exercising a little more caution than they would when assessing information
from the “trusted user-base” of the Cycology site. Interestingly, when asked about the issue of
Findings, Phase 2 – Role of social-psychological factors
trust and perceived reliability of the information on the website, many people criticised
themselves gently for being trusting, naïve or gullible.
Some people also implied that the unofficial nature of the site made Cycology more trustworthy
than an „official‟ website might be (e.g. information from a transport company or governmental
body) : “Not a corporate site if you see what I mean so I trusted it more” (Sarah). This
corresponded with findings in the exploratory research that word-of-mouth was in some
circumstances considered to be more trust-worthy than „official‟ information. Contrasting user
information from other cyclists with official cycling information provided by the city council, one
participant commented “the council always seems, in my mind, to have ulterior motives, to get
you to go a certain way, and maybe to ignore the fact that their other roads have potholes”
The previous paragraphs have demonstrated the high degree of trust which participants placed in
the group as whole, and this connects with the concept of group identification, which will be
discussed in the next section. It was clear from the participant interviews that trust in the group
as a whole was a stronger feature than trust in particular individuals within the group. This
contrasts with the findings from the exploratory research which showed that trust in word-ofmouth information was more likely to be strengthened by one-to-one similarities with the other
person than by a sense of shared group identity of any form (although there were indications from
the Phase 1 analysis that cycling might provide an exception to this). This underlines the point
that participant accounts in Phase 1 usually referred to word-of-mouth influences within dyadic
relationships, whereas Cycology created a specific group context in which to observe such
Within the Cycology project, clues provided and assumptions made about the identities of specific
individuals were reported to have made little difference to participants‟ evaluation of the
information posted. A straightforward explanation for this is that the interactions on the website
were not of sufficient quantity or length for participants to reveal a great deal about themselves,
although it might also be the case that the sense of trust in the group as a whole negated the
need to assess the credibility of specific individuals within it. However, despite this, it did appear
that some judgements were being made by some participants about other people‟s cycling styles
and levels of fitness, which could colour the way in which certain information was evaluated.
These factors are encapsulated within the final category depicted in Figure 10-2: „perceived
attributes of individuals‟.
Findings, Phase 2 – Role of social-psychological factors
Perceived attributes of individuals
These kind of judgements tended to be formed by comparing other peoples‟ comments about, for
example, the difficulty of getting up a particular hill, with an individual‟s own experience of the
same hill. If this suggested that the writer‟s level of fitness did not correspond with that of the
reader, then this would be borne in mind when reading other comments by the same person. This
matter will be discussed further in Section 10.4 on social judgement.
“ if you know a road that somebody has commented on, it's just easier to gauge if you
agree with their level. You know, what I think would be a nice thing, would be to have a
little rating, like "I'm an experienced cyclist, or I'm a beginner". You know, another way
of gauging what level somebody is at”.(Rachel)
Another participant remarked that he had got the impression that a person he interacted with on
the site was fitter than he was, and this slightly shaped his view of that person‟s comments:
“Obviously it helped that I knew what they were talking about and where they were
talking about. And it would make me investigate what they've said, without actually
completely believing what they said. Like they said, going up Pierces Hill, it's a bit of a
hill, but not a problem, sort of thing. Well, you're obviously fitter than I am, mate,
because (laughs) it is a problem!” (Mike)
For one participant, the ability to assess someone‟s level of fitness and compare it with his own
was something he would have liked to be able to gauge from the website, but was unable to do:
“you can make a better judgement (when) you can relate to the comments that are being
made, (…).. well it's not a case of saying "I don't believe you." Because I'm sure people
are saying what they believe is true. For them.(…) I took what people said on here in
good faith. But without the extra knowing them (…), without that extra bit of information,
I don't know how to relate that to my abilities”.(John)
However, another participant thought that judgements such as this about cycling styles or fitness
made no difference to the level of trust he would place in this type of cycling information. For
him, the degree of trust would reflect the other person‟s depth of experience and no more:
Phil: “So I know that there's quite a big variation in the type of cyclist. But it doesn't
really bother me. You've got the occasional cyclist and a racing cyclist using the same
Interviewer: Yeah, so it wouldn't affect how reliable you regarded what they'd written?
Findings, Phase 2 – Role of social-psychological factors
Phil: No. I think the only difference would be if someone was commuting every single
day and someone was commuting once a month. Then that person cycling every day
has got a broader depth of experience”.
10.1.3 Comparison with trust in other (formal) sources of information
Finally, participants were also asked to compare the trust they had in the route information on the
Cycology site with the routes generated by an automated cycling route planner such as the one
available on www.transportdirect.info. Consistent with the points previously discussed, many
people thought that they would trust the information on a user-generated site more because it
came from „real people‟ based on their experience, provided information about small but
significant details, and was known to be up to date, although many also pointed out the practical
advantages of a route planner over the Cycology site, such as speed of acquiring information and
greater breadth of information. One person said, comparing a route planner he had used with a
“social site” (as he termed Cycology):
“I think I still trust social sites more. Just because I don't think, I don't know enough
about the information they (i.e. route planners) use to plan it.” (Rick)
Comparing the two sources of information, another participant expressed a strong preference for
user-generated information, once again drawing on the dominant discourse within participant
accounts about the „real‟ nature of the information:
“It's real. I like the real. If you go to the AA, and get a route map, they'll send you in a
direction, and even if you look at it and think "well that's wrong" or "that's rubbish". But if
it comes from a person like this, you know somebody‟s actually tried it. And it just
makes it more real life. To me”. (Esther)
However, it was stated by many that a route planner has the advantage of being immediate, and
providing directions from the precise start point and endpoint required. Hence, both types of
route information fulfil different functions. Many people said they would look on a route planner
first, and then seek personal comments from users about the routes provided; the majority view
was that an ideal system would combine both forms of information (i.e. both formal and
informal). This suggestion will be discussed in Section 11.2 on desirable „social design‟ features
for advanced traveller information.
Findings, Phase 2 – Role of social-psychological factors
10.1.4 Informational, normative and referent social influence
Returning briefly to the theoretical constructs outlined in the opening paragraphs of this section,
we now consider how the trust factors described above might be conceptualised as channels of
normative, informational or referent informational influence (Deutsch and Gerard, 1955; Turner
et al., 1987). The calculus-based trust factors on the left of Figure 10-2 (intrinsic quality of the
information and comparison with own experience) can be linked to accepting others‟ advice as
evidence of reality, so influence ensuing from them might be categorised as straightforward
informational influence. The relational trust factors to the right (assumptions about the
benevolent motives of the group and perceived similarity with group members) incorporate a
social identity dimension; trust is enhanced through positive expectations about the reliability of
a reference group of fellow cyclists: hence referent social influence may ensue. Normative social
influence, in the manner defined by Deutsch and Gerard (1955) (i.e. complying with others to
seek approval and acceptance within the group), provides an unconvincing explanation, on its
own, of the mechanisms of trust and influence within the Cycology case-study, but would seem,
nonetheless, to be an intrinsic element of several trust factors. For example, the concept of
„reputation building‟ within the group (appearing in the upper right of Figure 10-2) implied a
normative pressure to provide trustworthy information and to be regarded, as one participant
articulated, “as a trusted member of the community”. The different conceptualisations of the
reference group within which these mechanisms of trust and social influence were played out will
be discussed in the following section.
10.2 Group identification („community‟)
This section will explore conceptualisations of the group (the „in-group‟ or „reference group‟)
within the Cycology project, and addresses that part of research question 4.4 which pertains to
shared membership of a community. The existence of one or more salient reference groups with
Cycology might be said to underpin all the social-psychological processes explored in this
chapter, and to have created a setting which was conducive to processes of referent
informational influence, manifested in the attitudinal and behavioural effects described in
Chapter 9.
Although expressions of group identification could be observed in some of the website postings,
most of the data in this section were generated from the interviews, and, to a more limited
extent, the questionnaire responses. Phrases such as community-building, cycling community
and virtual community arose without prompting in many of the interviews and questionnaire
responses, and these were interpreted within a framework of social identity and selfcategorisation theory (e.g. Tajfel, 1982; Tajfel and Turner, 1986, Turner et al., 1987). The word
Findings, Phase 2 – Role of social-psychological factors
„community‟ appeared twelve times within the questionnaires and 144 times in the interviews.
Three types of „community‟ were identified within participant accounts: a community of cyclists
generally; a work-based community; and a community of people within the project (identified as
either cyclists, co-workers or research volunteers) . Self-categorisation theory posits that group
identification contributes to cooperation (in this case, the sharing of information) and allows
referent informational influence to occur within the group (Turner et al., 1987). As Figure 10-3
suggests, these different communities were often felt to overlap, although they were ascribed
different levels of importance – or salience - by different people. For some participants, cycling
was the common factor which generated a sense of association within the project, whereas for
others, a greater sense of community arose from the knowledge that the participants worked for
the same organisation, or a small group of neighbouring organisations (a matter to be discussed
in Section 10.3 on effects of group size). Those participants who did not feel themselves to be a
member of a project community still made general observations about „the cycling community‟.
Moreover, self-categorisation theory holds that people‟s social identities are fluid, as they identify
more strongly first with one, and then another social group as the saliency of different group
memberships changes. Two factors which contributed to the sense of community in all three
forms were a sense of solidarity and empathy with other in-group members.
In this section, group identification within each of these articulations of community will be
discussed in turn, as well as two forms of emotional support associated with all of them:
solidarity and empathy (illustrated by particular postings). Finally, consideration will also be given
to those who did not experience a sense of group identification within the project, along with an
interpretation of the reasons for this. This gives rise to reflection on group identification factors
which might affect the „success‟ of virtual communities of travellers, which will be discussed in
Chapter 11 in the context of wider implications for advanced traveller information systems.
10.2.1 A community of cyclists
In terms of „the cycling community‟, it was generally felt that people who cycle are somewhat
more likely to band together and support each other than users of other transport modes.
Elaborating on why she believed the project participants to be “like-minded”, one person
expressed her sense of belonging to a general “cycling fraternity”:
“I suppose like-minded in the sense of people who cycle. It doesn't necessarily need to
be academics who cycle, or people who work at UWE who cycle, but more the general
cycling fraternity. Who have a…. I suppose some are more militant than others….. but
who have a view about promoting cycling - that it‟s a good thing to do.” (Sally)
Findings, Phase 2 – Role of social-psychological factors
Figure 10-3: Interlocking concepts of „community‟ in Cycology
Interlocking concepts of “community” in
Cyclists within
the project
Cyclists within
own workplace
A sense of group identification was expressed by another:
“I definitely feel that, being a cyclist, I definitely feel more of a community link with them
somehow, because I know they‟re cyclists”. (Kate)
It was suggested by some that this may be because cyclists consider themselves a vulnerable
minority which needs to work together to improve conditions for all cyclists:
“And so you've become, you create this sort of "us against them" mentality, just to keep
yourself safe. So everybody then clamours together. Because of power in numbers and
everything.” (Jess)
The high level, group categorisation of „all cyclists‟ was more likely than the workplace or project
categorisations to generate a sense of intergroup contrast with the users of other transport
modes (especially motorists). A fundamental premise of social identity theory is that positive
evaluation of the „ingroup‟ (which is required in order to improve self-concept) requires contrasts
to be made with a more negatively evaluated „outgroup‟ (Tajfel and Turner, 1986). Social
identity was found by Gatersleben and Haddad (2010) to affect cyclists‟ perceptions of other
Findings, Phase 2 – Role of social-psychological factors
cyclists – believing other cyclists to be like themselves. A sense of „cycling in-group‟ versus „cardriving outgroup‟ could be detected in some of the website posts, and was especially clear within
the interviews. The former is exemplified by the following posts on the website:
Never had any problems in the morning. At afternoon rush hour, full of
nutters, including one of two people who just hate cyclists.
Note: I‘ve been screamed at by motorists a few times along Kellaway Avenue.
No doubt for various injustices I have caused them. (Andy)
One participant considered the “us and them” situation to be an inevitable result of cyclists
feeling “beaten down” by motorists, but he also considered this to be regrettable:
“I think cyclists have a tendency to form quite a tight group, because there's a real "them
and us" type of mentality to it, which I find very annoying, as I drive as well (….)
Because there doesn't have to be this antagonism between cars and cyclists. But (…) if
I meet someone for the first time and find out they cycle, I end up talking to them about
cycling quite a bit. Whereas I'm not particularly interested in cycling, I don't know a great
deal about it, but you always end up with your war stories about cycling and horrors
about getting run off the road by maniac drivers.. And that sort of thing. So yeah, I think
it's a minority group that feel sort of slightly beaten down a lot of the time, you know.”
This concurs with qualitative research carried out among road users in four locations in the UK
by Musselwhite et al. (2010), which attributes the „us‟ and „them‟ focus of road users to the
highly competitive nature of road space in the UK. This previous participant statement also
exemplifies the fluidity of self-categorisations (Turner et al., 1987); people can move in and out
of different groups as their salience changes. In the example above, the participant describes
how his sense of antipathy towards motorists increases when he discusses cycling with another
cyclist (that is, when membership of a cycling ingroup is salient), but he is not entirely
comfortable with this, as he also drives a car on other occasions (implying that there may be
times when membership of a car-driving ingroup might have more salience for him).
The feeling of being in a minority, as well as a belief that more people should cycle, could also
make people more inclined to help one another, including sharing information. The greater
motivation of cyclists to share information and support, compared with users of other modes,
was particularly contrasted with car drivers, partly because sharing knowledge of a route such as
a “cut-through” rarely creates congestion among cyclists as it could well among car drivers:
Findings, Phase 2 – Role of social-psychological factors
“….cyclists sharing information doesn't deny them use of it. It's like, would the car
drivers put on excellent rat runs? Or no, because they want to keep their rat run to
themselves. Whereas cyclists have no problem telling you their rat runs, because
there's not enough cyclists to make it….. I've never been in a cycle traffic jam, if you
know what I mean”. (Mike)
However, it was also suggested that people who walk or use the bus may also experience a
sense of community, particularly the latter, who may have to share the “ordeals” of bus travel (to
which we return in Chapter 11.
10.2.2 A workplace community
In addition to the sense of identification with other cyclists, a feeling of association with others at
their workplace was apparent in the accounts of some participants. For example, one person
said she would have used the website in exactly the same way if it had been open to the whole
of her organisation because of a feeling that she had a “link with people”:
“I quite like the work thing, you know, it's that….. it gives you that connection to people if you
work at a place, or you're all here at UWE, you've got that link with people”. (Rachel)
Others articulated a sense of group identification not with others in their workplace generally, but
with those who commuted by bicycle in their company or organisation. The feeling of belonging
to a work-based cycling community appeared to be especially strong amongst those participants
who worked for the smaller organisations, all of which were considered to have a reasonably
strong cycling culture. They also commented that a lot of word-of-mouth information was shared
face-to-face, for example, at the bike racks or in the changing rooms. All of them also had an
email bicycle user group, although they were not thought to be used for the type of social
information-sharing which occurred on Cycology.
Some participants also mentioned what they thought to be the potential of capitalizing on a
workplace identity in order to encourage people to contribute to a website such as Cycology.
One person said that, if the website were available to all within a workplace:
“I think that there would be more of a sense of exclusivity and community between users,
with them more likely to input into the site” . (Alice)
Findings, Phase 2 – Role of social-psychological factors
10.2.3 A project community
Alternatively, a sense of community also appeared to be linked to the project itself, or through a
sense of connection to other people within it, either as a specific group of „fellow cyclists‟, or as a
group of people who had volunteered for the project. One participant remarked upon a feeling of
community within the project engendered by the custom of people answering other people‟s
questions. This was also evidenced by the way in which comments (particularly “gripes”) were
often quickly reinforced by other participants.
“…..because when I cycle in it‟s often quite early, and I don't see many other cyclists. So
it's quite nice. You get a sense of community when you see that other people are riding
exactly the same route, and most of the comments, they're annoyed about the same sort of
things that annoy you, or think that the same things that you think would also be a good
idea”. (Phil)
Interestingly, this participant implied that a sense of community within the project could only
ensue if there were something linking participants in the world outside, such as living or working
in the same area.
One participant described Cycology as a “virtual community of people who cycle which provides
a space for collaboration and help on all things cycling”. The degree to which Cycology might
accurately be described as an „online community‟ will be returned to in Section 11.4.
Expanding on this in her interview, she said:
“I guess part of that is that we have this common element of all working around the same
area. And that there were a limited number of participants. So I started recognizing
people‟s names. And that recognition made it feel like there's some kind of community
feeling to it. That it's just these 23 people, or however many were participating. So I knew
it was just "this is us, there's nobody else. This is our community. We‟re all cycling.”
However, not everyone shared this view. For example, one person felt disappointed that the
“community aspect” did not develop, due to insufficient interaction. Another thought Cycology
had the potential to develop into a community, but:
“You know, there wasn't a strong sense of, I didn't feel a sense of community like I do with
my housemates or with my other friends. But there was certainly a feeling of mutual
interest, shall we say. There were people there who were doing the same things as I was
doing, and were willing to chat about it”.(Adam)
Findings, Phase 2 – Role of social-psychological factors
Many participants felt connected to others within the project not only as fellow cyclists, but also
as a group of people who had volunteered for a research project. This also linked to their high
levels of trust in the information posted. One person thought the project engendered an inherent
sense of community which transcended other connections such as shared workplace or
“I mean, I feel part of UWE, I don't feel part of the MoD or Hewlett-Packard or anything,
so I wouldn't say I felt a sense of community with either of those, and I don't think the
geography either, to be honest. But I think the project gave it a sense of community. I
think it transcended where you work or anything….” (Julie)
This section has so far explored three broad concepts of community experienced in Cycology as
a whole; we now turn to two social factors inherent in some of the postings which were thought
to enhance these group identification effects, and which can be summarised as solidarity and
empathy. Fourteen people (in 25 references) mentioned the effect of a certain type of post
(such as „cycling in the rain‟, highlighted in Section 8.2.2) as something which contributed to a
sense of community amongst group members or cyclists in general, and this was often linked to
other concepts such as „mutual support‟, as well as shared experiences and emotions.
10.2.4 Mutual support (solidarity)
The idea of „solidarity‟ was raised spontaneously by a number of participants with reference to
specific posts. One participant remarked: “I enjoyed posts that encouraged a feeling of solidarity
amongst cyclists when people discussed annoying issues or comical aspects of a route”. One
such example from the website was entitled “Gripe with Pedestrians”:
Does anyone else ever find when crossing here on a bike that pedestrians
insist on walking in front of you and taking up the whole of the crossing
section? Grrr... particularly in term time... (Laura)
Yes, whenever I've cycled along this section I find that you have to
continually switch lanes to avoid pedestrians. Perhaps a few more logos
indicating space for cyclists and space for pedestrians would work? (Sue)
For others, this was a general impression they gained from the project as a whole. One person
said that the site “made me feel “solidarity” with the others who cycle and a kind of supporting/ed
feeling” (female): Another commented:
“I mean, I kind of got a sense that everybody doing it, you know, we‟re all cycling, everyone
Findings, Phase 2 – Role of social-psychological factors
kind of had an attitude of, you know, being willing to share information, help each other. It
was a nice feeling of solidarity in a way (…). So it was quite a nice feeling of community,
even though we were actually geographically not linked.” (Kate)
Solidarity is also associated with pro-social behaviour (Bierhoff, 2001), thus connecting it with an
area which will be discussed later in Section 10.5. There were also some overlaps between the
idea of Cycology providing social support for those new to cycling, and the way in which some
postings encouraged community-building (although practical information was perhaps
considered to be even more important). Was there a relationship between the amount of cycling
experience an individual had and the importance they attributed to „community/social support‟?
One person who had cycled to work for many years did not think the “cycling in the rain” type of
comment (quoted in the qualitative analysis of website posts in Section 8.2) prompted a sense of
community for him personally, as he had been cycling too long, but thought it might do so for
novice cyclists. Running two queries in NVivo showed that community building was indeed
mentioned most by people who had been cycling to work between 6 months and 2 years, and
the three categories of community identified previously were mentioned most by people who had
been cycling to work for between 1 and 2 years. Participants who had been cycling to work for
many years implied that social support for cycling was not something they particularly required,
because cycling was simply part of their routine; hence attitudes (and intentions and behaviours)
were likely to be more stable and less likely to be influenced by others. In social identity terms, it
might simply be the case that people‟s „cyclist social identity‟ becomes less salient as it becomes
a more habitual transport choice.
10.2.5 Shared experiences and emotions (empathy)
For some, creating a community was said to involve sharing feelings, motivations and
experiences. Others used terms such as “shared experiences”, “empathy”, “group feeling” and
“camaraderie”. These aspects appear to give the information greater resonance than if it were
merely instrumental.
“I think that, you know, if you're creating a sense of community, it‟s not only the
information that is getting across, but also the feelings and motivations. You know, "I
had a good day, I had a bad day." And those sort of shared experiences. That make a
sense of community. If it's just sort of "I go from point A to point B this way ", it's not
nearly as…, it doesn't touch you as much.” (Jess)
Findings, Phase 2 – Role of social-psychological factors
The community-building effects of some of the posts seemed to be especially strong when
participants were involved in interactions. The most active user of the Cycology website
remarked in her questionnaire:
“When people responded to my comments I did feel quite excited about being involved in the
cycling community, and was therefore encouraged to write more (…). Participation in this
project made me feel part of the cycling community which was quite nice. When I felt bad
about it, e.g. in rubbish weather, I knew there were others who had gone through the same,
which encouraged me to keep cycling- I am now an “all-weather cyclist”!” (Alice)
Like the concept of solidarity, empathy is another factor which is strongly associated with prosocial behaviour; the existence of both these factors may therefore have enhanced participants‟
willingness to contribute to the website.
10.2.6 Lack of group identification
Subtle distinctions among different participants‟ expressions of in-group/out-group contrast could
be detected within the Cycology project group itself. Two people explained that they felt
excluded from the group of participants, and possibly did not identify with a „cycling community‟
either (although one of these participants said that she felt more friendly towards cyclists than
other road-users because she believed that they ”looked out for each other”). With reference to
the theory of Turner et al.(e.g. 1987, 1994) these two participants did not categorise themselves
as members of the Cycology group. One of them, who was not cycling to work (although she
was contemplating doing so work), felt excluded because she believed that most other
participants were confident cyclists or had been cycling for a while. This inhibited her
participation on the site. She thought she would have felt more involved if there had been others
in the same situation, and if there had been communication between them:
“It seemed like, if there were other people who were only thinking about cycling, they
weren't really commenting or asking questions. So I felt a little bit like "oh, I'm going to be
the only one..". (…) Yeah, I think I was a little bit like "everybody else is already cycling.
I'm the only one who is not cycling. I don't really have that much to say." (….) I think it
would have worked out better if there'd been more people who weren't cycling”. (Helen)
The other participant who felt isolated from the group was the only person who was actively
negative about the project. Asked whether he thought postings such as “cycling in the rain”
evoked any feeling of „togetherness‟, he was the only person to say no:
Findings, Phase 2 – Role of social-psychological factors
Andy: “No. Because these are so, for me, these little posts, they were, you know, pretty
anonymous. Even though she is (she? –yeah) she was making an attempt to show some
personal experience, I don't feel any connection to her”.
Interviewer: “Right. Why was that then? Why didn't you feel any connection to her?”
Andy (laughing): “I don't have any idea who she is! Yeah. I mean, I guess I don't feel that
connected to the cycling community. That I feel like I should join in some general banter
about cycling experiences”.
Another interesting finding was that membership of the „project community‟, or at least one‟s
status as a trusted member, could be perceived as quite fragile. One participant thought that a
response to one of his routes, a response which he experienced as a criticism, might lead other
participants to consider all his posts to be unreliable:
“And it's a perfectly good route in my eyes. So after that I probably would think, what‟s the
point? Because you know, I'm now no longer a trusted user of this community.” (Adam)
Two people described themselves as having a rather unsociable personality which meant that
they did not „need‟ to feel part of a group. While believing that Cycology could engender a sense
of community, one of the two said that this was not important for her:
“But then, I think it's probably helpful for people who like to feel part of a group. To me it's
not important to know that other people are cycling. I'm cycling because I want to. I'm not
bothered if other people….. I mean, obviously, for the sake of the environment, I wish other
people were cycling, but I don't need to feel part of a group. But that's me” (laughs).
Using a framework of social identity and self-categorisation theory, this section has explored
several different senses of „community‟ which emerged from participants‟ accounts of the
Cycology project, as well as the way in which certain posts on the website conveyed a feeling of
solidarity and empathy. The next section will elaborate further on the group context by exploring
ways in which the limited size and composition of the Cycology group affected participants‟
experience of using the website.
10.3 Effects of limited group size and composition
During interviews and in their questionnaires most participants reported on what they considered
to be the advantages and disadvantages of the small size of the user group within the Cycology
project, and the possibility of recognising or being recognised by others because of the
Findings, Phase 2 – Role of social-psychological factors
workplace context. Most of the advantages can be summarised as social (for example, greater
trust in the information and community-building amongst group members, associated with
familiarity with other users), whereas the disadvantages might be categorised as instrumental
(membership not always large or diverse enough to provide a sufficient pool of knowledge or
basis for interaction). Observation of the website activity also supported this broad conclusion.
10.3.1 Positive effects
Starting with the perceived advantages of a limited group size and composition, it was reported
that the small size of the group (23 participants) created a sense of „intimacy‟, as it meant that
people could quickly start to recognize other participants‟ names (although it was rare for anyone
to gain a sense of the „person behind the name‟ unless they had happened to come across one
another in another context - see section on social judgement below). Because participants had
been recruited from a small number of organisations, there was a reasonable chance that they
would have met other participants „in the flesh‟ and recognised them on the website (aided by
the fact that many people used their real names when they wrote comments). However, it should
be noted that none of the participants knew that they knew anyone else when the project started,
and it was rare for anyone to report later that they had recognised more than one other person
whom they had come across in „real life‟. Unsurprisingly, the participants who most frequently
reported a sense of knowing who others were, all worked in one of the smaller organisations.
For most, this sense of social proximity provided an incentive for contributing to the site,
although two people reported that they had felt inhibited in what and how much they wrote
because they did not like the sense that they might be identified by colleagues (especially their
The small size of the group was thought by some to have increased their inclination, or sense of
commitment, to contributing to the website.
“Because it was small, it encouraged me to use it more. Because it was contributing to
something”. (Alice)
Asked if she would have contributed as much if the group had been bigger, one participant said:
“Probably yes, but I don't think I would have put as much time into it. Partly because
there'd be more people involved in it, and partly because it becomes more anonymous
the bigger the group gets (…).but it was a nice feeling, knowing it was a smallish group
and we were all doing it for the same reason”. (Kate)
Findings, Phase 2 – Role of social-psychological factors
“The same reason” refers to the fact that Cycology was a research project for which people had
actively volunteered to participate.
The motivating effects of limited group-size and social proximity with other participants concurs
with the outcomes of the EcoTeams approach to encouraging environmentally sustainable
behaviour at the local level (Staats et al., 2004). EcoTeams involves groups of householders
coming together to discuss their resource use, and agreeing to make changes to their behaviour,
which they monitor together. Parallels can be drawn between identified „success factors‟ in the
EcoTeams method (which produced significant resource savings, sustained over time), such as
the focus on practical information and tacit knowledge, and a supportive social element which
meant that people were more likely to carry out their good intentions if they made plans in front
of others (Avineri and Goodwin, 2010).
As discussed in Section 10.1 on trust, the small size of the group was also thought by some to
have improved the reliability of information on the website, and this may have partly been a
consequence of the sense of familiarity with other contributors:
“I think the small group aspect is certainly interesting, because after a while you come to
trust posts by people. Like I said with Marion, you think it's going to be amusing or
something quite nice about the route, and Alice will post stuff which is either very useful,
or she'll post something about your bike being stolen or something like that”. (Adam)
As previously discussed, the workplace context of the study meant that participants sometimes
recognised colleagues, or thought that they might be recognised, and this was thought to have a
moderating effect on what people wrote, ensuring that they provided accurate and reliable
“And I wouldn't really consider that someone would maliciously put bad information,
particularly in such a small….. maybe if this was open to the whole (…)…. I mean, it's a
small group, probably some people can figure out who other people are. Like I thought I
could figure out who a couple of other people were…” (Helen)
This finding is supported by several areas of social-psychological theory and empirical research
concerning people‟s greater propensity to behave in a cooperative or pro-social manner within
groups where they can communicate and where „dissenting‟ behaviour would be visible. This
was observed in social dilemma experiments conducted within a framework of Kelley and
Thibaut‟s Interdependence Theory (1978), which were outlined in Section 2.2 of the literature
review. Such experiments demonstrated that pro-social behaviour is more likely to occur within
Findings, Phase 2 – Role of social-psychological factors
small groups than in larger ones due to enhanced feelings of group identification, personal
responsibility, and ability to „make a difference‟ (Van Lange and De Dreu, 2001). Secondly, one
of the founding hypotheses of self-categorisation theory (Turner et al., 1987) is that factors such
as social interaction, similarity and reduced social distance all contribute to intragroup
However, claims in this area should be balanced against the observation that most of the
Cycology participants thought that reliability of (instrumental) information would not be a problem
on an open website either (i.e. in a much larger and more anonymous group context), albeit
they might treat information from anonymous sources on a public website with a little more
It was thought by some that a further advantage to limiting the size and composition of the group
was that it would be less intimidating for novice cyclists. Even if participants did not actually
know anyone else in the group, they did not regard the other individuals as „complete strangers‟
because many (at least in the case of the university participants) shared the same place of work
or study:
“I think someone who was thinking about cycling would feel more comfortable with it
being initially just UWE based (…). I think if someone was thinking this was for UWE
people to look at, you feel like you've got some association with other people. I suppose
it's just how you feel, isn't it.” (Ben)
10.3.2 Negative effects
Turning now to the disadvantages of the small group size, it was thought by most participants
that the group could have been bigger than it was in Cycology, but still retain its „small group
feel‟, and that this would have been beneficial in terms of increasing the reservoir of knowledge
within the group and also increasing the number of interactions, but without necessarily losing
the sense of familiarity among group members. Most participants thought that if a similar website
were to be created for general use (as opposed to a research project) either the university
population alone, or the university plus some of the smaller near-by employers, would provide a
sufficient reservoir of users, working on an assumption that the number of core users would be
only a small percentage of the overall user-base. In terms of ideal group size, the majority view
was that the group should be small enough for participants to have a sense of association with
one another, but to be drawn from a population diverse enough to provide both a wide
knowledge base and a degree of anonymity for those who wanted it (although some participants
Findings, Phase 2 – Role of social-psychological factors
disliked the anonymity and would have preferred to be interacting only with colleagues). More
specific estimates of ideal group size will be provided in Section 11.4.
Asked to compare, hypothetically, a website with restricted access, such as one within the
workplace, with an open access website, many participants saw advantages in the greater pool
of knowledge available on an open site. Some provided examples of when it would have been
useful for them to access user cycle routes outside their locality, including when visiting other
cities. Some also saw advantages in sharing information as widely as possible so that the
associated benefits could also be extended to as many people as possible. However, many
participants were concerned that an open access website could lead to a deluge of information,
making such a website confusing and difficult to use, and losing some of the social benefits
emanating from a smaller and more restricted user group.
This section had highlighted some of the effects of limiting the number of participants to 23 and
recruiting them from a small number of organisations (which was undertaken for carefully
considered methodological reasons, to which we return in Section 10.7 – reflections on the
methodology). In summary, these effects were see as broadly advantageous in psychological
terms (trust, group identification, accountability) and disadvantageous in a more practical sense
(reservoir of knowledge not great enough for all questions to be answered). In Chapter 11
consideration will be given to ways in which the advantages of „intimacy‟ (and usability) of a
restricted access website and small user group might be combined with the benefits of a much
broader knowledge base offered by an open group. For now, however, we remain with the
operation of the Cycology project; in the next section we move from consideration of factors
affecting the group as a whole to a more detailed analysis of some of the interpersonal
relationships within the group.
10.4 Social judgement
A matter explored at interview, following up an area of interest arising from Phase 1, was
whether any clues provided and assumptions made about the other participants‟ identities made
any difference to individuals‟ evaluation of the information. This section therefore addresses that
part of research question 4.4 pertaining to similarity and familiarity between information-giver
and receiver.
There seemed to be little interest in protecting identities in this small group environment: the
majority of participants used their real name (either first name only or full name) to sign off their
postings. Four female participants used only an initial and surname, or only initials, therefore not
indicating their gender, and a further female participant used a female-sounding pseudonym.
Findings, Phase 2 – Role of social-psychological factors
Eight uploaded a thumbnail picture of which four were portraits. Most people thought that the
use by some people of thumbnail photographs or avatars made the website feel more friendly
and personal, and allowed the reader to gain a very basic impression of that person. However, it
seemed that this was not enough to motivate more than a minority to upload a portrait of
10.4.1 Judgements about the group
As discussed in Section 10.1, most participants expressed the view that the individual
characteristics of the other contributors were not apparent and did not matter. The main
explanations given for the failure to obtain an impression of specific individuals were that the
project had not been of sufficient duration and the posts not numerous or long enough to allow
this. Hence, the judgements being made tended to be about the group as a whole, rather than
individuals (for example, that everyone was friendly).
Interestingly, many people assumed that everyone else in the project was a more
experienced/competent/serious cyclist than they were. Many were self-deprecating about their
own cycling competence. For some, the idea of the „serious cyclist‟ conflated with the
impression that other participants were „expert‟ or „activist‟ cyclists. As discussed in Section 10.2,
this reduced the motivation of some people (those who considered themselves particularly
inexperienced or „not serious‟) to contribute to the website, and in some cases made them feel
excluded from the group. However, this impression is incongruent with the observation that four
of the five most frequent contributors to the website had been cycling to work for less than two
years (and two of this group had started cycling to work within the past six months), and did not
appear to regard themselves as confident or especially experienced cyclists.
In addition to these generalised impressions about the group as a whole, some participants did,
in the course of the interview, reveal assumptions that they had made about particular
individuals - for example, interviewees usually used the male pronoun when referring to specific
individuals if their thumbnail photo or user name did not reveal gender. When questioned about
this, these interviewees could not explain why they assumed other participants to be male,
except possibly that they tended to see more men than women cycling (in fact, there were more
women than men on the project). It seemed that people were using their knowledge of the real
world as a yardstick to judge people on the website, rather than gaining an impression of
website users from their contributions per se. Just as people tended to assume that other
participants were male, because they tended to see more men cycling, one woman also said
she might assume that a man writing on the site was fitter than she was because the men she
sees cycling tend to go faster. Commenting that she knew more women than men who cycled
Findings, Phase 2 – Role of social-psychological factors
amongst her work colleagues, one participant thought that her impressions of male cyclists were
rather generalised (as fit and fast), and this might have extended to her impression of people on
the Cycology site:
“I suppose I'm also basing my judgement on the people I know from the faculty who
cycle already. And I don't really know the guys. The guys I see on my route are
whizzing by in their Lycra (laughs). You go "okay, you know what you're doing a bit
So I suppose it's my assumptions based on what I already see or know. In that way.
Rather than actual people I meet on the site. Does that make sense?”. (Rachel)
According to social-psychological theory, the process of overlooking individuality in this way
could be described as one where participants were drawing on stereotypes: shared beliefs about
personality traits and behaviours of group members (Fiedler and Bless, 2001). Within the field of
social cognition it is posited that social judgements are only partly determined by the stimuli of a
given situation, and also depend heavily on prior knowledge drawn from memory. People often
lack the time, cognitive capacity or motivation to consider all relevant information (or indeed,
insufficient information might be available from the given situation, as in Cycology). The lower
the capacity, motivation or level of new incoming information, the stronger the impact of prior
knowledge will be on the inferences and judgements made in a given situation. „Top-down
processing‟ occurs in this way when information-processing is driven by abstract, super-ordinate
knowledge drawn from memory, as the quotation from the previous paragraph suggests.
Conversely, „bottom-up processing‟ occurs when information processing is driven by new
incoming stimuli. Hence, the greater the depth of information-processing, the less influential prior
knowledge will be (Fiedler and Bless, 2001). Thus, a social cognition explanation of the
judgements made in Cycology about fellow group members would be that processing was
largely a „top-down‟ matter. One of the reasons for this may have been that the main focus of
attention was the content of the route information itself, so greater processing capacity was
devoted to the routes than the people suggesting them.
10.4.2 Judgements about individuals
Although most judgements were made about the group as a whole, it was reported in Section
10.1.2 that participants sometimes compared individual contributors‟ attributes with their own,
especially fitness levels, or confidence with regard to cycling in traffic, and used this to gauge
how well that person‟s experience might correspond with their own (especially ability to get up
steep hills). In this case, a greater degree of bottom-up processing may have been occurring, as
Findings, Phase 2 – Role of social-psychological factors
these participants were assessing cues about other people‟s attributes from the website
contributions themselves, before merging this with their own prior knowledge in relation to their
own attributes and experience. Thus, judgements being made about an informant‟s particular
attributes could affect the credibility with which the information would be received. This did not
equate to „not trusting‟ a person - quite the opposite, as one might trust a person to tell the truth
about their own experiences, without necessarily expecting that they would experience the same
things in the same way. Similar findings had emerged from the Phase 1 interviews in relation to
travel experiences and preferences more generally.
There was also a suggestion that the same process of comparing oneself with the other person
in the interaction with regard to, for example, cycling style or fitness, and making allowances for
this, occurs when information and advice is being given, as well as when it is being received.
One participant believed that the fact of not knowing who people were as individuals might have
inhibited her from giving detailed advice, for example about negotiating car traffic on busy roads,
in case the advice was not appropriate to the reader‟s level of confidence in cycling:
“But personally I don't like giving that kind of advice, because I don't feel…… because I
don't know the person. You know, if you knew that person…. then I probably would, but
I'd have to be very explicit about it, you know: "this is how I do it". But I'm quite happy to
have cars that close to me, or you know, take that risk myself. Whereas they might not
be. So I'd feel quite bad giving them some advice, which could end up not being such
good advice in the end”.(Rachel)
Finally, judgements were sometimes also made about other participants‟ attitudes towards
cycling – as far as was possible from the short interactions on the website. This is connected to
the assumptions, described above, which were made about other participants being „serious
cyclists‟. The concept of „serious cyclist‟ implies not only a certain way of cycling, but also a
certain attitude to cycling. One person said that he was happy to join in short discussions with
one particular contributor because he or she appeared to have a similar attitude to cycling to his
own, despite not being able to deduce very much about their character from such short
interactions (he assumed that this person was male when in fact she was female):
“You can't say someone‟s similar to you from that. But you know, they have the same
attitude to cycling, you know, they focus on… they weren't afraid to post little things that
might seem silly, but to say something that was typical of cycling, but didn't need to be
that grand or anything like that ”. (Adam)
Findings, Phase 2 – Role of social-psychological factors
10.5 Pro-social behaviour and attitudes
In this section we explore some of the motivations expressed by the participants for contributing,
or not contributing to the website, with a broad theoretical framework of pro-social behaviour.
This addresses research question 4.4.2, which asks why people are willing (or not) to offer
information to others.
10.5.1 Reasons for contributing
Participation in Cycology was voluntary, and everyone who posted on the website could be said
to have acted „pro-socially‟ in its broadest sense – that is, helping behaviour intended to improve
the situation of the help recipient, which is not motivated by professional obligation, and which
may be either egoistically or altruistically motivated, or a mixture of the two (Bierhoff, 2002). In
so far as voluntary information-sharing was the raison-d‟être of Cycology, pro-social behaviour
was in some ways an essential component of the whole project. In addition to the posting of
route information, evidence of pro-social behaviour could be found in the warnings about new
problems along cycle routes, such as: broken glass on a roundabout; recurrent danger points on
routes; the notification of a local cycling event, which concluded with the remark “I wouldn‟t want
anyone to miss out”; and the way in which many (although not all) questions posted on the
website were answered. Not all the posts might be described as helpful in a practical sense,
especially those which involved „sharing a gripe‟ about poor infrastructure or the behaviour of
other road users, although even these were construed by some as psychologically helpful in
terms of community-building and boosting cyclists‟ morale in the face of adversity.
Three levels of helping could be detected in the observed behaviour of the Cycology
participants: first and foremost providing helpful information to other participants; secondly,
assisting the researcher and the project as a whole; and thirdly, at a more abstract level,
supporting a transport choice for short trips (i.e. cycling) which may be construed as more
beneficial to society as a whole than other transport modes. The three levels are depicted in
Figure 10-4, which also indicates the associations of the different types of helping with „hard‟ and
„soft‟ rewards; this will be discussed later in this section. Thus, at the first level, pro-social
behaviour took the form of posting information which was helpful to other participants in a
practical way, or making observations which might enhance others‟ enjoyment of cycling and
reinforce a sense of „community spirit‟. For some participants, writing this kind of post was
motivated not only by wanting to be helpful to other participants, but also the researcher, and/or
to make the project „a success‟ because they supported the idea of sharing cycling information in
principle. Posting questions on the site could also be interpreted as a pro-social act at both the
Findings, Phase 2 – Role of social-psychological factors
Figure 10-4 Hierarchy of pro-social behaviour in Cycology
first and second levels, in that it helped to stimulate interaction on the site for the benefit of other
users and the researcher, even if a straight-forwardly egoistic motive was also present: that of
seeking information for the questioner‟s own benefit. At the highest and most abstract level,
engaging in an activity which might play a part in encouraging cycling (which all participants
believed Cycology did, or had the potential to do) could also be regarded as a pro-social act in
relation to wider society because of the reputation of cycling as a pro-environmental transport
choice (Tapp et al., 2010). Of these three levels of pro-social behaviour, the first is of most
interest for the present research, and is therefore the focus of this section. The extent to which
some participants were motivated by wishing the project to be a „success‟ (second level) is also
addressed here, although this is also a methodological issue with implications for research
validity, to which we will return in Section 10.7 (reflections on the methodology). The third level the extent to which cycling might be regarded as a „pro-social transport choice‟, although a
related topic of interest, did not lie within the scope of the research questions and is therefore
excluded from the present analysis.
The main reason given by participants for contributing to the site was that it was a matter of “give
and take”. Most people said they wanted to be helpful and “share stories”, but also hoped that
they might learn something useful for themselves, especially alternative routes and details about
them which were unavailable from other sources. One of the most frequent contributors summed
up her reasons for posting simply as “It seemed like a friendly, helpful thing to do”. She later
mentioned the appeal of “getting something back”, but for her, this was not the main issue:
Findings, Phase 2 – Role of social-psychological factors
“You know, it's nice to have a bit of something back. And as you noticed, I posed the
question about the cycle path without much response. But no, I just sort figured, if
there's something that might be useful to people, it‟s not very difficult to share it”.(Sally)
The comment that it was not very difficult to share information suggests that, for this person at
least, the „pro-social actions‟ were not very costly to the help-giver.
Motives of reciprocity were also demonstrated by the following:
“And I felt that I had something to add in telling people about my route and just passing
on information. So that was really why I did it, and I was quite interested in whether
anyone else had a route”. (Rick)
“Just to contribute something really, just to put something up there. It would have been
interesting if there was somebody doing a similar commute to me, you know if someone
had had an alternative route which would have served me, it would have been
interesting to see”. (Doug)
The interest in obtaining feedback as well as providing a route which might be of use to others
was exemplified by this early posting accompanying a route drawn on the website:
If anyone knows a better route - please let me know!!! It's not too bad
apart from the bits around kellaway avenue/coldharbour road where the
roads are narrow and quite steep, and the same going up to the downs on
Clifton Down road. (Rick)
This participant said in interview that he was especially interested in hearing other people‟s
views about his route, suggesting that a self-serving motive was at least partly responsible for
telling others about his route. Two people responded in this thread, one confirming that she
believed this to be the best route, and another suggesting an alternative route, which the original
contributor then tried and reported back on. Thus, this interchange might be interpreted as prosocial on the part of all three participants, with motives which might be both altruistic and
A number of theoretical approaches might be drawn upon to help explain this reciprocal helping
behaviour: social exchange theory (Homans, 1958), reciprocal altruism (Trivers, 1971; Fehr and
Fischbacher, 2003) or as evidence of the combining of an altruistic and egoistic motive system
(Bierhoff, 2002). The present findings support Bierhoff‟s (2002) assertion that “in many real life
Findings, Phase 2 – Role of social-psychological factors
examples, a mixture of altruistic and egoistic goals seem to motivate pro-social behaviour”
(p.331). Of the different explanations of pro-social behaviour which were outlined in the literature
review (biological, individualistic, interpersonal and social-systems theories), the present
research particularly supports a social systems approach because there appeared to be a norm
of helpful behaviour within the „project community‟. The social-systems explanation of pro-social
behaviour emphasises the influence of factors inherent within a social system, such as cultural
norms and rituals shared within a community and argues that people are socialized into a
particular pattern of behaviour which reflects the „rules‟ and expectations within their community
(Bierhoff, 2001).
For some, the reciprocal „taking‟ in return for „giving‟ took an affective form as well as, or instead
of, a practical form. A third of the participants articulated a feeling that it was “nice to help
people”. One person explained how pleased he felt to know that his route had been helpful to
someone, noting that he was surprised to realise this as he had assumed that his motives for
going on the website were confined to the more practical matter of improving his knowledge of
“I thought that was nice. It was just the fact that I had done a route and I'd put it on there
and someone else benefitted from that. I really liked that actually. It made me feel good
to be honest, the fact that I‟d put the route on there and someone else was benefiting
from it. It's nice to know that. I initially went on there to see if I could find out whether
there were other routes that I could come across, so I didn't expect that to begin with.
But it was nice to know that perhaps you'd saved someone some time in the
Ben also described this feeling as "a bit of an ego boost".
The most active contributor to the website reported feeling pleased when others responded to her
comments, and this encouraged her to write more. For her, the positive feelings she experienced
as a result of contributing were linked to a desire to feel part of “the cycling community”:
“I felt quite good about posting comments, although I was quite aware that some of them
may have seemed quite trivial, especially as at the time of this study I was quite new to
cycling and had not experienced much. When people responded to my comments I did
feel quite excited about being involved in the cycling community, and was therefore
encouraged to write more”.(Alice)
Some participants said that if they liked a particular route themselves, they tried to make it sound
interesting and pleasant in order to encourage others to try to it. Just one person made a direct
Findings, Phase 2 – Role of social-psychological factors
reference to positive self-presentation as one of the rewards for acting pro-socially by offering
useful information:
“When it asked for a username or a pseudonym, I always put a user name. On a site
where you‟re posting, I want people to know it‟s me. If I'm posting a route I want to take
some of the glory for that”. (Adam)
Rewards such as enhanced reputation and personal satisfaction are termed „soft rewards‟ by
Hall and Graham (2004) in their study of motivations for contributing to an online group called
CipherChallenge, through which code-breaking enthusiasts cooperated to solve a puzzle. They
identify these rewards as important alongside the “hard or explicit rewards” of access to
information and knowledge. The sense of "collective morality" based on reciprocity which
developed within the group studied by Hall and Graham was echoed within Cycology. These
authors caution against labelling such reciprocal helping as „altruistic‟ due to the existence of soft
rewards, but this reciprocation is consistent with the broader term „pro-social behaviour‟.
Another „soft reward‟ for contributing to the Cycology site was the sense that this might be a way
of encouraging people to start or keep on cycling. Half the participants revealed in their
interviews that they liked to encourage others to cycle whenever the opportunity arose (although
they did not identify themselves with the image of the “evangelical cyclist”). This contrasted with
the exploratory research, when most participants said they were very reluctant to try to influence
other people‟s travel choices, suggesting that people reveal themselves to be more „directive‟
when faced with a specific situation, than they would admit to be when discussing travel
behaviour more generally. Alternatively, this may be a specific „cycling effect‟; people who cycle
may be more concerned than users of other modes to convince others to join them.
In addition to helping others and seeking practical (hard) or psychological (soft) rewards from
other participants, some people participated actively and regularly because this was a timelimited project which they wished to be successful - thereby exhibiting pro-social behaviour
towards not only the other participants, but also the researcher. One participant, who had hoped
for more interaction on the website, admitted that he was making a conscious effort to contribute
because he wanted the project to work well, as well as wanting to obtain more benefit
“Yeah, I mean, I was conscious, to a certain extent, after a while I got to the point of
looking for things that I could respond to. Because, partly from the perspective of
wanting it to work as an experiment, and partly because I wasn't getting much from it, I
Findings, Phase 2 – Role of social-psychological factors
thought maybe if I put more in, then I might get more back. So I did make a conscious
effort to try and respond to posts”.(John)
The evidence that some participants were behaving pro-socially towards the researcher as well
as other participants within the confines of the project demonstrates one of the biases which
result from studying an artificial setting rather than a natural one (for example, an existing online
community). This matter will be discussed further in Section 10.7. However, when probed on this
matter, most of the group expressed the view that they would have behaved in a similar way on
the website if it had been a permanent feature within their organisation or locality, and not a
research experiment. Although this cannot be demonstrated without further research in a „natural
setting‟, it might be hypothesised that those who would contribute to a „real-world‟ website might
have similar pro-social motives to those who would volunteer for a research project of this type,
since the same positive approach to information-sharing about cycling and a desire to make the
initiative work might be in place. Evidence of this can be found within the literature on behaviour
within real world online communities; for example – as previously mentioned - Hall and Graham
(2004) found evidence of “ good citizenship” and reciprocity within the Yahoo group
This section has so far discussed expressions of pro-social behaviour in the form of
contributions to the website, as well as participants‟ motives for making these contributions. Most
of the discussion so far has, therefore, related to the more active contributors. However, nearly
one third of the participants wrote only one or two messages, and it is also of interest to
understand why they did not contribute more.
10.5.2 Reasons for not contributing
An emerging online community such as Cycology requires a regular flow of contributions in
order to survive and cannot function if members are interested only in reading others‟ postings referred to in online circles as “lurking” (see Kozinets‟ definition of an online community as a
"group of people who share social interaction” (2010, p.10)). Indeed, many online services built
on user-generated content, such as www.tripadviser.com or the movie reviews website studied
by Ling et al. (2005) require a critical mass of user reviews in order to be able to operate, and
these authors have applied much thought to how people might be encouraged to contribute
more. Studies of online communities have shown that it is common for a small group of active
members to contribute a disproportionately large share of the messages (e.g. Ling et al., 2005;
Zhang and Watts, 2008). For example, in the online travel (backpacker) discussion forum
studied by Zhang and Watts (2008), the most active 2% of members posted 25% of all the
messages. Adar and Huberman (2000) found that on the peer-to-peer music sharing service,
Findings, Phase 2 – Role of social-psychological factors
Gnutella, 10% of users provided 87% of the music. This phenomenon was also found in
Cycology – albeit to a lesser degree - where the most active 5 participants (22%) wrote 55% of
the posts. If posting routes, comments and questions was essential to the project, and
represents a pro-social behaviour, as we have we argued, which was both egoistically as well as
altruistically motivated, why did the quieter members behave in a manner which was „less prosocial‟?
The most common reason cited by those writing low numbers of posts (two or less) were: lack of
relevant knowledge (relating to lack of, or limited experience of cycling to work, or only ever
cycling the same route); lack of time; and lack of interest. This is summarised in Table 10-1. As
will be discussed in the section on typologies, there was not always a direct link between
enthusiasm for the project or cycling knowledge (expressed in interviews and questionnaires)
and level of engagement with the project (reading and writing posts). Among the seven
participants who wrote two posts or fewer, two were unenthusiastic about the project mainly
because they rarely or never cycled to work, which meant that they believed they had little
knowledge to share:
“I was happy to post information. But, as my cycling is limited to getting to UWE, I didn‟t
know what else to post after I presented the route I take”.(Andy)
Another of the low contributors was, however, a frequent and long-standing commuter cyclist
who felt that he had little to learn from or offer the project because he was cycling the same way
every day from an area where none of the other participants lived:
“I did not really think I could post anything of any great interest to start a thread, but I did add
to an existing thread a couple of times. If I‟d had anything relevant to say I would have
posted more often”.(Jim)
This serves as a timely reminder of the obvious point that pro-social behaviour in an online
community is unlikely to be stimulated if there is perceived to be no direct relevance to the
interests of the user, with regard to either seeking or offering information.
Three more of the seven low contributors were enthusiastic about the project but felt they had
limited knowledge to offer. Asked how she felt about posting, one of this three said: “bit scared
as didn‟t wanna seem an idiot”, demonstrating self-presentation concerns. Several of the low
contributors reported later that they had been reluctant to post because they believed that all the
other participants were more knowledgeable about cycling than they were (although this was not
in fact the case).
Findings, Phase 2 – Role of social-psychological factors
Finally, one of the seven was both enthusiastic and experienced but said he had had little time to
contribute during the project period, due to work pressures. As evidence of his interest in the
Cycology concept, this participant reported that he had later recommended the cycling pages of
the main bristolstreets website widely in his company, after several hundred new employees
began work at his offices over the summer following a company merger.
Table 10-1: Stated reasons for low level of posts (seven lowest contributors)
little knowledge of
little enthusiasm for
not enough time to
These three types of constraint were not limited to those who posted the fewest times. In
particular, many people (including some of those who contributed the most) said that they did
not have the time to log in and post on the website as much as they intended. As most
participants used the website whilst at work, time constraints were usually attributed to work
pressure, although some participants also went away for short periods of the project. In addition,
some found certain technical aspects of using the site, such as the log-in procedure to be timeconsuming and a disincentive to writing posts .
Two of the more frequent contributors were actually rather unenthusiastic about the project, but
had taken an active part due to a sense of obligation, suggesting that they were exhibiting
aspects of a “pro-social personality”, which is associated with high levels of personal
responsibility (Bierhoff, 2002). This meant that they acted pro-socially despite facing the same
constraints which inhibited others from contributing.
This section has demonstrated how the majority of participants acted pro-socially in the
Cycology project, and did so at one or more of three levels (helping other participants, helping
the researcher, and helping to encourage a modal choice construed as pro-social). An analysis
was also conducted of the reasons why one third of the participants contributed little (perceiving
themselves to have little knowledge, little enthusiasm for the project, or insufficient time).
Findings, Phase 2 – Role of social-psychological factors
This part of the chapter has elaborated on five interlocking social-psychological factors which
help to explain the psychological and behavioural effects of both informal and formal
information, shared through word-of-mouth within the case-study information system. Before
concluding this chapter, we now briefly consider some of the „non-social‟ factors influencing the
use and effects of the website information, which provide an important context for the socialpsychological mechanisms previously discussed.
The role of other („non-social‟) factors
10.6.1 Instrumental factors
In Section 5.4.4 (role of instrumental factors in Phase 1 data) it was noted that instrumental
factors, such as availability of alternative transport services, have a major influence on the
context of information-use, including whether or not travel information is sought or used in the
first place. This was also illustrated in the case-study, where – unsurprisingly - the practical
usefulness of information was seen as essential to the project (indeed, this was understood to
be its main purpose). Within Cycology, „usefulness‟ generally referred to geographical relevance:
geographical information had to be relevant to the places people were going in order to be of
practical use. Most often, people were interested in alternative routes between their homes and
place of work. As one person summed up, the website was “Very useful if you happen to cycle
those areas, less so if you live/work elsewhere or don‟t yet cycle that far” (Helen).
Although some participants read comments regardless of their geographical relevance,
sometimes because they were “fun” or interesting to read despite relating to an area through
which they did not usually cycle, the majority referred in their questionnaires to a preference for
reading posts that were geographically relevant to them, and therefore useful. Most of this group
tried out a new route provided by another participant, although one person, for whom no relevant
new routes were posted, said that the website had given her the confidence to work out new
routes for herself due to the usability of the map itself and the general sense of encouragement
she obtained from belonging to the group. Interestingly, in this case, social as well as
instrumental factors had contributed to practical benefit. Conversely, five participants said in their
questionnaires that the project was not useful to them because no alternative routes (i.e.
previously unknown to them) for their commute were suggested. One of this group stressed that
without information of practical use, the project had no influence on what he did:
“Umm, I mean, I've been cycling, doing that commute for quite a long time, and I guess
I'm fairly settled in my habits as to what I do, and I didn't see anything that would influence
what I did. There were some interesting comments, but there was nothing I took away as
Findings, Phase 2 – Role of social-psychological factors
being directly usable, or especially useful for me”.(Doug)
Here, a direct link was being made between the instrumental role of the information and
influence on behaviour in terms of route choice, suggesting that the social mechanisms involved
in information-exchange (discussed in this chapter) may have been more effective in influencing
attitudes and general motivations for cycling rather than having a direct impact on route choice.
This quotation also suggests a link between length of cycling experience, habitual behaviour,
and lack of interest in, or susceptibility to the more social aspects of the website. Unsurprisingly,
those who found the project to be of little practical use were, in the main, those who had been
cycling for a long time, and tended always to use the same route to work because there was little
alternative, or because none had been suggested. However, most of these participants said that
they would have found it especially helpful when they had just started cycling, or were new to
Bristol or their place of work, or when they wanted to cycle to another new place.
Similarly, another participant, when asked if she personally had gained anything useful from the
interaction on the site, said:
“ No, but then it depends on "do you need it or not?". And I didn't feel, apart from, as I
say, when I asked the question "is this road still shut?", "yes, it is" - that was very helpful.
If you have a specific need. Because again, I have my route, I'm happy on it. And some
people are looking for example, for bike buddies, and it would be very useful for that sort
of thing. If you tend to be not interested in that, and you don't feel you need it, you
wouldn't use the site in that way”.(Elaine)
Those who were primarily interested in the instrumental purpose of the site said they did not,
personally, value it as a social forum. As one person commented, “I‟d use it as an information
source, rather than a social interaction site”. This participant (Mike) did indeed only contribute to
interactions about his route to work, although he employed a personal and friendly style on the
website (rather than an impersonal, pragmatic one), suggesting that friendly social interaction
was not an end in itself but a means of achieving the main purpose – the practical business of
sharing instrumental information.
Although this section has emphasised the importance of the direct instrumental role of the
Cycology information, and focussed on the views of the sub-group of participants who saw it
chiefly in these terms, many participants expressed the view that social types of information (for
example, shared experiences and emotions) could also be „useful‟ in a different way. As the
preceding sections have shown, social information was thought to play a significant role in
building up a sense of community and mutual support, which could, in turn, encourage people to
Findings, Phase 2 – Role of social-psychological factors
cycle more, as well as simply making the site more enjoyable and fun to use. In this sense,
social interactions could be seen as playing an indirectly instrumental role in encouraging
general cycling behaviour via the reinforcement of pro-cycling attitudes and intentions. Again, it
was thought that information-sharing, as in Cycology, could be particularly useful, in both
senses, for those who were new to cycling or to a particular workplace.
Another participant highlighted a temporal dimension to the instrumental and social roles of the
website. He remarked that he had first seen the website as a source of (route) information, but
as time went on and everyone had posted their routes, the community dimension became
“At the start it was for the information aspect, the informative aspect, but towards the
end it became for the community aspect”.(Adam)
This perception is supported by the changes in the content of website postings presented in
Figure 8-5, which showed that route information peaked it the first week, whereas other types of
comment remained more steady over time.
Having discussed instrumental factors in the sense of the practical usefulness of the Cycology
information, we now give brief consideration to the inevitable impact of broader instrumental
factors in the „outside world‟ on people‟s willingness to change their behaviour in response to
new information. This was the sense in which instrumental factors were considered in Section
5.4.4. It would seem obvious that word-of-mouth information about cycling, even if more
persuasive than information from official sources, will not on its own be sufficient to start people
cycling or cause them to cycle more if they face practical constraints. Because all but one of the
case-study participants were already cycling to work at least occasionally, few examples of such
personal constraints were provided in the interviews. However the participant who was not yet
cycling to work cited an instrumental reason - lack of sufficient fitness - as her main justification
for not cycling (even though she had felt encouraged, to a degree, by her involvement in
Cycology, and had learned of a suitable route). The participant who cycled to work the least
frequently cited lack of bicycle lights and lack of cycling skill as her main reasons for not cycling
more (at least in winter), although she also described a psychological barrier relating to the fact
that she had once witnessed a serious accident involving a cyclist.
10.6.2 Individual factors: personality and past experience
Although detailed consideration has been given the role of social factors in information-use
within this thesis, it is important, as suggested in the previous section, to remember that this is
Findings, Phase 2 – Role of social-psychological factors
only one part of a complex picture of information behaviour, in which individual factors such as
personality and past experience remain a major element. In this section we outline a number of
such individual factors affecting the giving, reading and using of information on the Cycology
Firstly, information behaviour within the case-study was always individual in that people were
looking at and writing the messages privately (often at work), and processing the information
cognitively; the physical use of the Cycology website was not a „group activity‟ in this sense.
Personality factors affected how people regarded the website - for example, whether they
enjoyed the social interaction - and how open they were prepared to be about themselves (selfdisclosure). As previously noted, two participants described themselves as “unsociable”,
explaining that they did not tend generally contribute to online forums unless there was a specific
information need, or to answer a direct question .
“I'm not very sociable anyway, as you know, so I tend to sort of the lurk on these sites. I
get the information I need, and think "that's a good idea, I'll bear that in mind". Or, "that's
a good lock" or whatever I'm worried about at the time. And act accordingly, you
A further personality-related factor affecting individuals‟ activity on the website was degree of
concern with self-presentation (Leary, 1995); expressed as a reluctance to demonstrate lack of
knowledge (for example, Sarah was a “bit scared (to post) as didn‟t wanna seem an idiot”). By
contrast, the most frequent contributor, Alice, reported feeling worried that her comments were
sometimes trivial, but wrote them anyway.
The influence of past, personal experience on the validity attributed to information acquired from
others has already been discussed with regard to trust (Section 10.1); for example, participants
spoke of comparing the comments about routes which they knew with their own experience. If
such comments coincided with their own opinions, other comments about unfamiliar routes
would also be seen as reliable. In some cases, evidence of disagreement based on personal
experience also appeared on the website.
For some individuals, their level of cycling experience affected the degree to which they
contributed to the website. Notably, some who considered themselves less experienced reported
that they had written few posts because they felt they had little knowledge to offer. However, this
was not a consistent finding. The most frequent contributor (Alice) had started cycling to work
only during the previous six months, and described herself as having little cycling experience.
Conversely, some participants who had been cycling for many years (e.g. Jim) contributed little.
Findings, Phase 2 – Role of social-psychological factors
The latter finding suggests that more experienced cyclists were less engaged with the website
because they had a lesser need for information, but this was not necessarily the case, as most
expressed themselves as being receptive to new ideas such as alternative routes. For example,
Rachel remarked that she had expected to participate in the project mainly as an informationprovider, but had learned more from it than anticipated. A more detailed picture of individual
factors which affected the use of the website by four participants (Elaine, Andy, Laura and Phil),
each representing a „user type‟, will be presented in 11.4.2.
10.7 Reflections on the Phase 2 methodology
The Phase 1 research had revealed some of the limitations (as well as the strengths) of the
interview method in exploring word-of-mouth information-use, and for this reason it was decided
that data from observation of actual behaviour should supplement reported beliefs, attitudes and
behaviour in Phase 2. The Cycology study was illuminating in what it revealed about the
difference in interpretation which may arise from different data sources. In some cases,
participants‟ own accounts - through questionnaires and interviews - of their engagement with
the website provided a different view from that which had been obtained through observation of
their actual activity on the website (i.e. logging in, opening markers and writing posts). The most
notable example of discrepancy concerned participants who logged onto the website
infrequently and wrote very few posts (e.g. Phil, Jess and Esther), creating an impression of
disengagement, but who expressed high levels of enthusiasm for the project during interview.
Some of the reasons for this discrepancy were discussed in 10.5.2. Conversely, other individuals
(e.g. Elaine and John) were more negative than their actual activity on the website had
suggested. Thus, the two data sources together allowed a more comprehensive analysis than
either would have done on its own.
An „artificial‟ research environment
A further methodological strength lay in the innovative nature of the case-study. Not only was the
Cycology platform technologically innovative, but also exemplified the relatively new
methodological area of creating „online spaces‟ in order to study particular facets of social life
(Section 2.5 of the literature review). This latter point, does, however, give rise to questions
about the effects which the nature of the project – i.e. an online environment set up specifically
for the research, as opposed to a „naturally occurring‟ online community – may have had on the
findings. To what extent, for example, was participants‟ behaviour affected by the artificial nature
of the research environment and the fact that they had volunteered to take part? Whilst it was
clear from the interviews that this did inevitably influence behaviour (e.g. some people were
behaving pro-socially because they wished to assist the research), most participants expressed
Findings, Phase 2 – Role of social-psychological factors
the view that they would have behaved in a broadly similar fashion if this had been a „real-world‟
website, albeit their participation might have been reduced in the absence of a sense of
commitment to a time-limited project. Clearly this remains untested, but participants‟ statements
in this respect are moderately convincing since many were enthusiastic about the principle of
information-sharing about cycling, and saw practical benefits for both themselves and the cycling
community, regardless of whether this was occurring in a „natural‟ or „artificial‟ setting.
A related matter for reflection is the degree to which the introductory briefings for participants,
held just prior to the start of the project, might have stimulated the sense of community which
arose so clearly in the findings. Might these face-to-face meetings among some of the
participants have interfered with, either positively or negatively, the process of group
identification? It is possible that this occurred for some; for example, two people who met at one
of the briefings exchanged contact details and later became friends. However, this development
did not appear to affect directly their use of the website, as they did not interact with one another
there; in fact, one did not realise who the other was (she used a pseudonym). Interestingly, one
person obtained the impression from the briefing he attended that the other participants were all
deeply committed cyclists (unlike him), and reported in the interview that this had given him a
negative attitude to the project from the outset. However, this was the only case where a direct
effect from the briefings arose in the data. Ultimately, the effects should not be exaggerated
because the briefings were attended by small numbers - one was attended by seven participants
and the other by six. Moreover, there was no direct association between, on the one hand,
attendance or non-attendance at a briefing, and on the other, level of contribution to the website
or degree of stated enthusiasm for it. However, the potential for prior face-to-face meetings to
bias the findings is a matter which deserves consideration if similar case-studies were to be
Reflections on epistemology
It is also apposite to return briefly to the epistemological issues which were discussed in Section
3.2.2 and consider the effects of the research process on the nature of the data. A reflexive
approach requires an appreciation that the data were generated (and not collected) in a
dynamic process through participants‟ interaction with the website, their interactions with one
another, and those between participants and the researcher. Although the descriptive analysis
reported in Chapter 8 emerged from a largely literal reading of numerical data, most Phase 2
data were qualitative and subject to an interpretative reading. A double hermeneutic (Smith and
Osborn, 2008) was created though the interpretations of both participants and researcher.
However, the findings were not construed as merely a product of the research process, nor a
social construction in the postmodern sense, but were closely connected to people‟s activity
Findings, Phase 2 – Role of social-psychological factors
outside the research (i. e. cycling to work in real life); ontologically, the existence of an external
reality to which the findings directly relate, is espoused.
Limitations in the data
At a more practical level, an assessment was made of whether the case-study had generated
sufficient quality and quantity of data to answer the research questions, and also whether
enough interaction had occurred on the Cycology website, and was of sufficient use to
participants, to be regarded by them as a successful pilot initiative. Overall, the number of posts
and log-ins to the website provided enough data to conduct a detailed analysis and, with the
important addition of the interviews and questionnaires, allowed answers to be found to the
research questions. However, with the benefit of hindsight, a larger number of participants
would have extended the breadth of information exchange and possibly led to greater number of
interactions on the website (although restricting Cycology to a group of under 25 allowed the
researcher to gain deeper knowledge of each individual than would have been possible with a
larger group, and contributed to a degree of familiarity within the group).
Sixty people originally volunteered for the project, of whom 30 were selected in accordance with
the criteria described in Section 7.6, with the expectation that 25 would actually participate.
Attrition was slightly higher than anticipated as a number of volunteers withdrew for a mixture of
personal and work-related reasons. A further reason, which was unexpected and only
discovered at the end of the project, was a technical one: the computer system at one of the
organisations from which participants were recruited did not support the website effectively. With
hindsight, a greater number of the original volunteers should have been invited to participate, to
allow for unexpected attrition. Moreover, the target number of final participants should, ideally
have been higher than 25, but perhaps no greater than 35, in order to maintain sound project
management and to allow the majority of participants to be interviewed. The „ideal‟ group size for
a „naturally occurring‟ community of online users (without the additional management and datacollection/management requirements of the PhD) is a matter of debate which will be discussed
in Section 11.4.
The quantity of data obtained from the website itself might also have been greater had the casestudy run for longer than 6 weeks (perhaps 8-10 weeks). This would have needed to have been
decided at the outset when participants were provided with documentation clarifying their
expected involvement. Extending the project whilst it was underway would have required the
agreement of all participants, and was, moreover, considered impractical for Cycology because
the main summer holiday period was about to begin.
Findings, Phase 2 – Role of social-psychological factors
Finally, it was noted in Section 8.1 that it was not possible to gain a complete understanding of
which of the website postings had been read by whom, and when, because the daily email
digest made it possible to read the messages without logging on to the website. Only when
participants were logged in was it possible for the system to record which of the website markers
were opened by whom. The email digest was intended to serve as a daily reminder of the
website‟s existence (at which it was reportedly very successful), but it was not anticipated that
participants might stop after reading the digest and not follow the links to the website itself. For
this analysis the researcher had to judge the veracity of participants‟ self-reports (in interview)
regarding their reading of the email digest, and such reports could provide, at best, only a broad
picture. This was not considered to be serious gap, as quantitative data were chiefly contextual
and formed only a minor part of the overall data set, but the observation of reading behaviour
would have been more complete if the gap had been filled. If a similar case-study were to be
designed, one way of collecting data on the reading of the email digest might be to request an
acknowledgement of receipt each time a participant opened a new email containing the digest.
Naturally, this would still not guarantee that an email had actually been read, nor would it
capture data on the reading of individual messages within the digest. Another option might be to
dispense with the digest altogether, but this would entail the loss of a deliberate similarity with an
internet discussion forum. Without an email reminder, participants might forget to check the
website at all, with serious consequences for the case-study overall, although a compromise
might be to include only part of each message in the email, necessitating a visit to the website to
read a complete post.
In conclusion, this was a voluntary project which requested a longer time commitment and
greater engagement from participants than that required by many research projects - and no
„hard‟ incentives such as cash were offered - yet 21 of the 23 agreed to be interviewed
afterwards and all offered constructive feedback on Cycology‟s strengths, weaknesses and
potential for future development, suggesting a degree of success both as a means of data
collection and a real-world information system.
Findings, Phase 2 – Role of social-psychological factors
10.8 Chapter Summary
This chapter examined the role of social-psychological factors in five areas which were
prominent within the analysis: trust, group identification, group size, social judgement and prosocial behaviour. Two groups of „non-social‟ factors were also outlined to provide additional
context: instrumental and individual factors.
Firstly, in the section on trust, it was explained that all participants believed the information
posted on the site to be reliable and trustworthy. This took the forms of both calculus-based trust
and relational trust, although the former was more important. Calculus-based trust arose from
the intrinsic quality of the information and through a comparison with participants‟ own
experience. Relational trust took the form of assumptions about both the good intentions of the
group as a whole, and perceived attributes of individuals. Secondly, several different senses of
„community‟ (group identification) emerged from participants‟ accounts of the Cycology project:
a community of cyclists generally; a community of cyclists within the project; a community of
people within the same or neighbouring places of work; the project was also described by some
as a „virtual community‟. Group identification was enhanced by the way in which certain posts on
the website conveyed a feeling of solidarity and empathy, and this was thought by some to
increase the resonance (impact) of the information. Group identification and related trust
mechanisms were found to have enhanced processes of referent social influence.
Thirdly, the effects of the limited size and composition of the Cycology group were examined.
Limiting the group size and composition had „social advantages‟ (for example, greater trust in the
information and community-building amongst group members, associated with familiarity with
other users), but „instrumental disadvantages‟ (membership not always large or diverse enough
to provide a sufficient pool of knowledge or basis for interaction). The fourth section explored
social judgements made with the group. Judgements tended to be based on attributes of the
group as a whole, but where judgements were made about individuals, they tended to be based
on stereotypes and reflect comparisons with people in „the real world‟ rather than actual
judgements of people using the website („top-down processing‟).
The fifth area explored was pro-social behaviour. Three levels of helping were identified: helping
others in the project (by contributing to the website); helping the researcher; and behaving prosocially towards society in general by using an environmentally friendly transport mode.
Reciprocal altruism („give and take‟) was identified as the main reason for pro-social behaviour,
as participants were found to have obtained „soft rewards‟ for helping others (e.g. enhanced
reputation and personal satisfaction), as well as „hard rewards‟ in the form of useful, practical
Findings, Phase 2 – Role of social-psychological factors
information. Three main reasons were provided by those who contributed little: sparse
knowledge; lack of enthusiasm for the website generally; and lack of time to contribute.
Finally, a number of points were made about the role of instrumental and individual factors
affecting word-of-mouth information-use. Within Cycology, „usefulness‟ generally referred to
geographical relevance: geographical information had to be relevant to the places people where
were going in order to be of practical use (direct instrumental role of the information), but many
participants expressed the view that social types of information (for example, shared
experiences and emotions) could also be „useful‟ in a different way. Social information played
an indirectly instrumental role in encouraging general cycling behaviour via the reinforcement of
pro-cycling attitudes and intentions. It was thought that information-sharing as in Cycology could
be particularly useful, in both senses, for those who were new to cycling or to a particular
workplace. With regard to individual factors, both personality-related factors (e.g. concern with
self-presentation or degree of „sociability‟) and past experience in cycling were found to affect
information behaviour.
This chapter completes the discussion of findings in response to research question 4. The next
chapter will now turn to research question 5.
Implications for ATIS
Chapter 11 : Implications for the wider field of advanced
traveller information
In this chapter the final research question of the thesis will be addressed:
How might information-sharing amongst travellers be applicable to advanced
traveller information systems more generally?
The following sub-questions were defined:
5.1 Could advanced traveller information systems which currently offer mainly „formal‟
types of information benefit from the integration of „social design features‟ (types of
user-generated content incorporated into Web 2.0 systems), and if so what specific
form could these take?
5.2 To which modes of transport (beyond cycling) might information-sharing be most
5.3 What features of a „user community‟ might be conducive to word-of-mouth
information-sharing (e.g. group size, composition, types of user).
The aim of the chapter is thus to provide additional discussion and draw broader conclusions
about the applicability of the findings from the exploratory research (Phase 1) and the casestudy (Phase 2) to the area of advanced traveller information. Answers to research questions 5.1
and 5.2, discussed in Sections 11.2 and 11.3 respectively, draw principally on a thematic
analysis of the interviews with Cycology participants, during which the strengths, weaknesses
and potential future development of an information system of this type were discussed. In
addition to the thematic analysis of interview and questionnaire data, research question 5.3 is
also addressed (in Section 11.4) with the assistance of the numerical analysis of website usage
reported in 8.3 and a non cross-sectional analysis of the different data sources (website
observation, interviews and questionnaires) relating to four individual participants, each
representing a “user type”. The technique of non cross-sectional qualitative analysis was
described in Section 7.9.2. As well as drawing on participant responses and researcher
interpretation, this chapter also incorporates a system developer‟s perspective obtained through
a post-project interview with Toby Lewis, the developer of www.bristolstreets.uk and the
Cycology layer of the website. The findings in this area thus represent a series of reflections on
potential implications of the empirical findings for advanced traveller information systems,
highlighting where further research or validation could bring further insights. Firstly, however,
Implications for ATIS
consideration is given to the principles of generalising from case-study research and the extent
to which broader claims can or cannot be made on the basis of the research carried out within
the Cycology study.
11.1 Generalising from the case-study
In the previous chapter, findings were reported from the study of informal information-sharing
amongst users of an innovative information system – a single case, in a specific location,
concerning a specific mode of transport. Before turning to the implications of the findings for
advanced traveller information systems more generally, consideration must now be given to the
principles of how far is it valid to generalise beyond the specific case, by returning briefly to the
methods literature. The problem of external validity has been a common critique of case-study
research, although some proponents challenge the very assumption that knowledge must be
generalisable in order to have value. As Flyvbjerg (2006) argues: "That knowledge cannot be
formally generalised does not mean that it cannot enter into the collective process of knowledge
accumulation in a given field or in a society." (p.227). Forms of generalisation appropriate to the
research in this thesis were introduced in Chapter 4.4, but in this section we consider those
which specifically concern case-study research.
Notwithstanding Flyvbjerg‟s position that “often it is not desirable to summarise and generalise
case studies” (2006, p.241), a degree of generalisability has been sought in this thesis in the
form of both an elaborated theory of information behaviour, and specific findings from the use of
the Cycology website which might be translated to other advanced traveller information system
contexts. Theoretical generalisation is typical of case-study research (Yin, 2009, Robson, 1993).
In this case, it was sought by generalising this particular set of findings to broader theory by
comparing the empirical results with, principally, the assumptions of, and findings related to
social-psychological theories of behaviour, as well as findings from the literature on (travel)
information-use, and behaviour within online communities. The translation of specific findings
from the Cycology study to other traveller information system contexts constitutes a form of
inferential generalisation (Ritchie and Lewis, 2003), and this was sought through comparisons
with the literature, supplemented by comparison with contemporary, real-world developments in
social media and transport.
In the domain of theory, case studies can be useful not only for testing existing theories in the
literature, but also for building new ones by testing new propositions or hypotheses (Flyvbjerg,
2006). In this thesis, existing models of information-use were elaborated upon, at the end of
Phase 1, through the addition of the social factors found to be active when travel information,
both formal and informal, is communicated through word-of-mouth. In Phase 2, these additions
Implications for ATIS
to existing theory were both further elaborated, and validated within a specific information
system context. Thus, one of the outputs of the research was a new model elucidating the role of
word-of-mouth information within travel behaviour (presented in 11.5). However, the theory must
be tested in other similar contexts and the same results replicated, if the theory is to gain strong
support. Yin (2009) refers to this as the “logic of replication”, arguing that “the theory (.…) that
led to a case-study in the first place is the same theory that will help identify the other cases to
which the results are generalisable.” (p.43). Thus, the model of word-of-mouth influences on
travel behaviour would now require testing in other contexts where travel information is
communicated within social networks (either through electronic or face-to-face word-of- mouth) –
but this must remain outside the scope of this thesis .
Whilst the behavioural model may be applicable in a broad range of contexts (although this will
require testing), the key aim of Research Question 5 was to seek generalisation in the specific
application area of advanced traveller information. It will be recalled that in Chapter 7 the
Cycology website (or „information system‟) was identified as the main unit of analysis of the
case-study, in order that generalisations might be made to other information systems (Yin,
2009). Cycology was categorised as an extreme or unique case (in relation to traveller
information), as no other system of precisely this nature had been identified. The logic of
generalisation from a unique case is that it can be used as an exemplar with which to compare
other cases as they arise (Yin, 2009, Flyvbjerg, 2006). In the intervening period since the
Cycology case-study was designed, technological developments in the field of social media have
continued apace, which means that at the time of writing, more applications have indeed arisen
in the field of traveller information (incorporating user-generated content) which could be directly
compared with Cycology. Whilst this means that the recommendations made in this chapter
may now appear less innovative, the evidence that similar developments are now actually
occurring adds legitimacy to the findings and might be interpreted as a form of validation for
some of the practical suggestions which emerged from the research.
11.2 Desirable „social design features‟
Today‟s traveller information systems are extremely diverse, offering journey-planning, real-time
travel information and navigational advice through a range of technological media. As the
suggestions proposed in this section cannot address all types of system, the category to be
considered here is web-based journey-planning and real-time information systems, accessible
via computer, mobile phone and other portable devices, as this is a technological medium which
offers considerable scope for social interaction amongst users, as well as being most directly
comparable with the Cycology website. This section seeks to answer research question 5.1:
Implications for ATIS
5.1 Could advanced traveller information systems which currently offer mainly „formal‟ types of
information benefit from the integration of „social design features‟ (types of user- generated
content incorporated into Web 2.0 systems), and if so what specific form could these take?
Transport has not been immune from the recent explosion in digital word-of-mouth, but – as
discussed previously in this thesis - there has so far been little convergence between the sharing
of informal information based on the personal experiences of transport users, and formal travel
information delivered by transport operators and/or government agencies through advanced
traveller information systems. However, the current research has shown that both forms of
information are considered highly complementary, and can, together be influential in the travel
decision process. This section considers some of the „social design features‟ incorporated into
the Cycology website which were viewed positively by participants, as well as their suggestions
for improvement. In considering these matters, the research incorporated a degree of „prototype
testing‟ by users of an innovative information system. Of particular interest were the suggestions
for incorporating elements of user-generated information into automated route planning systems.
Inevitably, such proposals represent what was considered by (non-technical) users to be
desirable, and not a discussion of what might be technically feasible, for which further research
would be required. Most of the recommendations in this area, therefore, led on directly from
participant responses during interviews, although in some cases researcher interpretation
formed part of the process of developing recommendations from participant responses. In
addition, a system developer‟s perspective was obtained from Toby Lewis, who was first asked
for his feedback on the functioning of the Cycology website prior to his being informed of
participants‟ opinions.
11.2.1 User recommendations
Many of the perceived benefits of the user-generated content on the Cycology website have
already been discussed in the previous chapter, as they were particularly closely associated with
the reliability which participants attributed to the information. Such benefits were often identified
by comparing and contrasting them with „formal‟ types and sources of information, such as
paper maps and online maps and route planners (such as Google maps), and often referred to
social aspects of the site. Benefits included the extra level of detail provided by „real users‟, the
ability of users to keep information up-to-date, and the email digest which helped to keep it fresh
in peoples‟ minds and enhance the interactive dimension.
Implications for ATIS
Another aspect which was appreciated was that posts were kept short and to the point, without
the „rambling‟ and sometimes negative nature of some of the postings which appear on cycling
forums or email groups (described by some participants as “ranting”). It is conjectured that this
could have resulted, at least in part, from the small size of the comments boxes on the website.
Although there was no limit to the amount of the text which could have been added, the small
boxes could have provided a cue to keep posts short. The ability to upload a thumbnail portrait
or icon which would automatically accompany an individual‟s posts (as with social network sites)
was also considered by participants to be a positive feature which was thought to „personalise‟
the site, although only a third of participants actually used this facility.
Participants identified the main disadvantages of the Cycology site compared with formal travel
information sources (such as automated route planners and static maps) as the time required to
obtain answers to route questions, and the limited geographical coverage, which meant that
some questions could not be answered at all. The „instant routes‟ provided by route planning
systems and their wide geographical remit meant that they were considered more attractive than
a site such as Cycology when routes were being sought at short notice or outside the immediate
locality. It was noted, however, that few participants knew of the existence of journey planners
specific to cycling, and none of those who did could recall the actual name of services such as
www.cyclestreets.net, or Transport Direct‟s cycle route planner.
A number of technical improvements to the Cycology website were suggested in the interviews.
Many thought a desirable feature was the ability to post photographs to illustrate the route
comments. In fact, this facility was available in principle, but suffered from technical problems
during the project period. Interestingly, only the two most enthusiastic and frequent participants
attempted to upload photographs to the site (and were unable to do so), which could suggest
that only a minority of very active users might take the time to upload photographs if a system
such as Cycology were to be developed further, due to the extra time required above and
beyond writing comments, as well as the additional inconvenience of stopping to take a
photograph during a cycle ride. However, this may be an area where norms might develop once
an idea has gained currency; for example, at the time of writing, over 25,000 photographs had
been uploaded by users around the UK to www.cyclestreets.net.
The popularity of the thumbnail portrait feature (in participant accounts if not actually in practice
on the site) has already been noted, but one participant went further by suggesting the addition
of tick boxes through which users might provide a little extra information about themselves
(similar to the brief personal profile information which one may provide on social networking
sites). A specific example would be the option to self-categorise oneself according to degree of
cycling experience. This might help information-seekers to assess how far they might expect to
Implications for ATIS
experience aspects of a route in a similar way to particular information-providers. As suggested
in Section 10.1.2 on relational trust, an experienced cyclist might be more likely to heed
warnings of heavy traffic and steep inclines as expressed by another experienced cyclist, than if
such concerns were voiced by a novice cyclist. The information would be more likely to be
considered subjectively valid if it emanated from a person thought to be similar. Within Cycology,
a sense that other participants had a similar level of „cycling competence‟ might also have
encouraged some of the less confident cyclists to contribute to the website more. In fact, there
were several people who considered themselves to be beginners, or not confident cyclists, or
believed that they lacked the knowledge to contribute actively. If they believed themselves to be
outside the norm within the group, they were likely to feel somewhat excluded, showing how
important it was for people to feel some sense of similarity and group identification with others in
order to engage. As discussed in detail in Section 10.2, this finding was illuminated by selfcategorisation theory (Turner et al., 1987). Interestingly, many people thought that everyone
else was more experienced at, or more serious in their approach to, cycling than they were. It is
conjectured that a „self-rating‟ feature (through which most participants may well have selfcategorised themselves as „non-serious‟ cyclists in some way) might have helped to dispel this
belief and encouraged the quieter participants to contribute more.
A further participant suggestion for helping users to assess the validity of the information was the
incorporation of a user rating system, such as the invitation to agree or disagree with the
statement “I found this review helpful” (or “I like/dislike this route”), commonly found on websites
offering user-generated reviews. As discussed in Chapter 2.5 and Chapter 8 of this thesis,
research (e.g. Cheung et al., 2009) has found user ratings of consumer recommendations to
increase the perceived credibility of online recommendations through a process of normative
influence. This should perhaps be balanced against the risk that some people might be inhibited
by the prospect of their posts or routes being „rated‟, especially in a small group. A rating system
might be more helpful on an open website, where posts are likely to be more anonymous.
A suggestion raised by a number of participants was linking the map to a more „standard‟ type of
discussion forum rather than relying on „floating comments‟ to discuss matters not linked to a
particular location (i.e. comments boxes which hovered over the map). It was thought that a
forum environment might be conducive to higher degrees of interaction on more general topics.
This slightly contradicts the view that the succinctness of the comments, possibly related to the
small comments boxes, was a positive feature of the website. Only through the testing of
different formats could it be gauged whether the length of posts was a feature of the website
format or simply the style of those particular users. The juxtaposition of maps and discussion
forums/email groups is already a feature of some local cycling campaign websites such as
Implications for ATIS
www.camdencyclists.org.uk and www.camcycle.org. At the time of writing, Camden Cycling
Campaign‟s online forum included an area specifically for route requests, but only three
messages had been posted. The community-based cycle route website www.cyclopath.org in
Minnesota, USA, had elicited 109 route-related comments in 28 threads in its linked discussion
forum between its introduction in April 2010 and February 2011. An advantage of a forum
structure over the Cycology website is that discussion topics can be found quickly, whilst on
Cycology it was necessary to scroll through floating comments chronologically, which may have
become awkward had the project continued for longer.
There were a number of suggestions as to how information on Cycology might be targeted to a
user‟s geographical area of interest. One such suggestion was that a user might elect to receive
within their email digest only those comments within particular postcode areas. This might be a
useful feature should a system attract a higher level of usage than was the case during the
project. It was also suggested that some messages, filtered by geographical area and/or topic,
might be sent out instantly to mobile phones, in a similar way to the short messages („tweets‟)
shared within Twitter groups. In this way it would be possible to warn other cyclists immediately
of current obstacles or hazards (such as the warning on Cycology of broken glass on a
roundabout) along their regular cycle routes. These might be analogous to the „official‟ travel
alerts currently delivered by many traveller information systems (such as Transport Direct and
Transport for London‟s online travel information) or indeed „tweets‟ sent by transport operators
about occurrences in the transport network. Another way of obtaining the most geographically
relevant information would be the capability to search for user routes between one postcode
area and another; once again, this might prove useful if the system were to attract so much route
information that a filtering system would be required to prevent the map from appearing overcrowded. The filtering information by geographical area is possible on the Minnesota cycle route
website www.cyclopath.org, where users may specify „watched regions‟ and elect to receive
email notification when changes are made to the map in those areas.
It was clear that the Cycology website fulfilled a role different from that of an online journey
planner, with the latter providing instant and comprehensive information but lacking the personal
dimension which made the user-generated routes appear to be more trustworthy. Rather than
preferring one type of online information over the other, most participants felt that the two were
complementary, and in an „ ideal world‟, both might be combined. It was suggested that, if
technically feasible, a route might be generated instantly by a journey planning facility, but user
comments referring to sections of the route could also be retrieved by the system and appear as
comments boxes on the map. There was a general view that it would be useful to be able to
comment on the routes generated by an automated planner, and to ask other cyclists what they
Implications for ATIS
thought of them. One template for incorporating user comments onto computer generated routes
has been developed for www.cyclopath.org , where sections of the route can be selected to
reveal „notes‟ added by users.
In general, user feedback is incorporated into route planners through dialogue between
individual users and website/system managers; for example feedback is actively encouraged by
www.cyclestreets.net. This allows route information to be updated and routes generated by the
system to be optimised, but through a process which is not visible to other users, suggesting that
the „social advantages‟ arising from a sense of identification with other users (such as greater
perceived reliability) are not incorporated. Interestingly, at the time of writing, funding is being
sought by the developers of CycleStreets to add a feature allowing cyclists to plot their own
cycle routes and contribute route comments (c.f. http://www.cyclestreets.net/routes/, accessed
To a certain degree this type of convergence is already occurring on the Camden Cycling
Campaign‟s online map, where routes contributed by users can actually be compared with
routes generated by www.cyclestreets.net on the same map. At the time of writing, however,
user comments on the Camden map do not, in general, provide the level of detail or the type of
personal interaction which emerged through the Cycology contributions. This may, in part, be a
function of the greater size of the Camden user base, the greater number of routes on the map,
and the open (publically accessible) nature of the website. It might also result from the fact that
the Camden website includes a discussion forum where interactions can take place (although,
as previously noted, little discussion of routes had taken place at the time of writing).
Finally in this section, a suggestion was made about how social proximity and group
identification within the small group might be maintained whilst extending the geographical
scope and overall user-base of the website. One way of achieving this might be to set up subgroups of users within an organisation or locality, under the umbrella of a larger, open system.
Users could choose whether their routes/comments were to be accessible to everyone, or just
members of their own group. Ideally, it would be possible to contact other subgroups for
information if knowledge from outside the immediate area were sought. This is analogous with
the different levels of access on social network sites such as Facebook, or the way in which
Liftshare operates using both an open system and restricted access groups within organisations.
In addition to the advantages of such an arrangement in terms of obtaining and offering
practical information, design features which allowed the formation of groups within groups might
also stimulate and build upon processes of group identification, which were found in Chapter 8 to
encourage referent social influence among users.
Implications for ATIS
11.2.2 A system developer‟s perspective
Following the analysis of participant interviews, the developer of www.bristolstreets.co.uk and
the Cycology layer, Toby Lewis, was interviewed in order to obtain his independent assessment
of the project, ascertain whether it had stimulated any ideas for further development in the field
of social media and traveller information, and discuss the recommendations arising from the
Lewis reported that the project had led him to consider developing the Cycology concept as a
service which he could offer commercially to travel planners in larger companies and
organisations in the local area. This would take the form of a map-based website where informal
travel information might be shared among employees about all local transport modes. As well as
its use for cycling, he believed that informal information-sharing could be particularly helpful for
bus travellers and for car-sharing – a view shared with many participants, which will be
discussed in the next section. Lewis outlined the idea to travel planners attending the city‟s
Green Commuter Club in late 2009, as well as the City Council‟s travel information team, from
whom the concept received a favourable response in principle; but it was thought that budgets
were unlikely to stretch to the purchasing of such a service in the present economic climate,
when many travel planning posts were disappearing. Therefore, he was dissuaded from
pursuing this as a viable business option for the time being, but believed it to be an idea worth
following up should economic conditions improve. He reported being interested in the apparent
social benefits which ensued from the closed environment of Cycology (as opposed to the
publically accessible bristolstreets website) and this had contributed to his view that it might be a
marketable idea of interest to travel planners. Lewis also predicted that the next big development
in online mapping would be in improved website compatibility with mobile devices (notably smart
phones), with the current market for „apps‟ (e.g. CycleStreets apps for iPhones and Android)
providing a transitionary phase.
Lewis also reported a number of design-related ideas which had emerged from the project,
many of which corresponded with suggestions which had been made separately by participants.
Notable among these were the use of more non-location-specific, social networking type
features. For example - using Facebook terminology - people might be given the option of
posting more personal profile information, and of selecting geographically (and socially) close
sub-groups with whom to interact more frequently than their whole „friend‟ network. Facebook
might even be used as a gateway interface through which „friends of friends‟ might be brought
into the network. Echoing the recommendation made by participants for a linked discussion
forum, he acknowledged that bubbles on the map might not necessarily be the best place to
Implications for ATIS
locate discussion, and that comments and responses might perhaps be better placed to the side
in the manner of a „normal‟ discussion forum. Each marker on the map could, for example be
linked to the forum rather than opening as a bubble. He regards it as a limiting factor of
www.bristolstreets.co.uk that people are not used to websites which are not driven by the text,
and may need a visual interface more simple than a map to which their eye is initially drawn
(rather like the chronological list of comments on the sidebar on Cycology).
This part of the chapter has provided a number of propositions for desirable technical features,
from a user perspective, which might be incorporated into an „ideal system‟ for sharing online
traveller information. These might be best suited to those systems where the map is a central
feature, such as www.walkit.com (for walking), the Google maps route planner (mainly driving or
walking), or the cycle route planners previously mentioned such as www.cyclestreets.net.
Transport Direct also offers a map view of its route plans, as well as the location of incidents on
the transport network. Most user recommendations corresponded with conclusions drawn
independently by the Cycology developer. Ideas for further „social design features‟ from both
these sources are summarised in Table 11-1.
Table 11-1: Recommended „social design features‟
User recommendations
Developer ideas
Combining user comments with routes
generated by automated planners.
Route planning algorithms which incorporate
user feedback.
Enabling user subgroups (interacting
amongst themselves) to develop within a
wider system.
Enabling users to „friend‟ geographically (and
socially close) sub-groups.
Allowing users to filter routes/comments by
geographical area.
Allowing users to filter routes/comments by
geographical area.
Receiving updates (including real-time) on
comments relating to specified geographical
Better mobile access via smart phones
(allowing, for example, users to receive
relevant updates whilst en route).
Linking a discussion forum to the map.
Putting comments to the side instead of on
the map, with every marker perhaps linked to
a forum rather than opening as a bubble.
Allowing users to specify personal profiles.
More social networking type features (e.g.
Facilitating the use of profile pictures and the
uploading of photographs to the map.
Technical features for more efficient
uploading of photos (e.g. automatic file
Linking to Facebook to draw in „friends‟.
Implications for ATIS
11.3 Potential relevance to other transport modes
This section addresses the following research question:
5.2 To which modes of transport (beyond cycling) might information-sharing be most
It reports on questionnaire responses and interview discussion about the potential usefulness of
web-based information-sharing for transport by modes other than cycling, the types of informal
information which might (or might not) be beneficial, and the extent to which users of other
modes might identify with one another and be willing to cooperate. Like the previous section, the
findings here relate to hypothetical scenarios, as no one had engaged in this form of online
information-sharing in relation to utility travel apart from the Cycology project.
A distinction was clearly drawn between the different potential uses of such a website for
different transport modes. Identified uses fell into four categories: route-finding; other items of
informal, practical information and advice; community-building amongst users; and drawing
issues to the attention of transport operators or local authorities. As the main focus of
information-sharing on the Cycology site had been routes, the most commonly suggested
alternative use was for walking, as route-planning was thought to be an important factor in
cycling and walking (and car driving – although motorists were already thought to be well served
by „conventional‟ route planners, satellite navigation systems and motorists‟ discussion forums).
The route-finding dimension was thought to be less relevant for public transport users, whose
routes are already fixed by the service provider – unless travellers have a choice between
different service alternatives. User comments could, for example, complement the walking
routes generated by systems such as www.walkit.com in the ways suggested for combining
cycle route planning with users‟ routes, which were discussed in the previous section.
Although the route-sharing element is less relevant for bus travel, other types of informal
information were thought by participants to have a helpful role to play in improving the
experience of local bus travel – particularly those which might reduce uncertainty. Bus travel was
a potential application area which was also indentified by the Cycology developer during the
follow-up interview. Several participants remarked that it would be useful to obtain from other
travellers the kind of detail which bus companies might be reluctant to provide – for example if
particular buses were consistently either early, late or over-crowded. Hence, bus route and
timetable information from the bus companies could be complemented by user comments about
the experience of using the service at different times, leading to an improvement in the breath of
public transport information available. For infrequent bus users (as the majority of Cycology
Implications for ATIS
participants were), online information-sharing was thought to be potentially useful for obtaining
up-to-date information such as changes to routes and road-layout (for example, whether the
introduction of bus lanes was speeding up particular journeys).
It was also suggested that this might be a useful means of conveying informal information about
the basic norms of bus travel for those who had not used this mode of travel for some time, and
were apprehensive about doing so; for example whether it was necessary to „flag down‟ a bus or
whether it would stop anyway - although in theory this type of information might also be provided
by the bus company, or by a combination of both providers and users. For example, the
Transdev Yellow bus company website includes detailed advice on using buses, including:
“please clearly put out your hand so the driver knows you wish to board” (www.bybus.co.uk,
accessed 15.2.11).
Most of the examples of „useful informal information‟ referred to local bus travel rather than other
public transport modes, although some also mentioned its potential use for train travel; buses
may have been given the greatest consideration because the context of the Cycology project
had been information-sharing about local commuting, amongst a group of people for whom
travelling to work by train was rarely a convenient option (many did not live close to a local train
The community-building dimension of information-sharing, which had been apparent within
Cycology, was also discussed in relation to other transport modes. Generally participants
believed that the sense of belonging to a community within which people support and encourage
one another (in-group identification, c.f. Tajfel, 1982, Turner et al., 1987) was likely to be
stronger among people who cycled than amongst users of other transport modes. This was
attributed by participants to a tendency for cyclists to consider themselves a beleaguered
minority in a transport culture dominated by the car, and perhaps because cyclists have a
stronger motivation to „convert‟ others (social identity amongst cyclists was discussed in Chapter
10.2; see also Gatersleben and Haddad, 2010; Skinner and Rosen, 2007). However a number of
people did suggest that a similar sense of group identification might develop amongst regular
bus users, particularly those using the same buses and routes and travelling to similar
“I've heard people say the brotherhood and sisterhood of cycling, which is a bit corny,
but there is that sort of a feeling. If you get a puncture by the side of the road,
someone‟s going to stop and give you a hand.(…) People who are trying cycling more
need to know what the new routes are, so I think (….) this community seems to work
well for cyclists, but I also think it might work well for public transport, say buses.” (Julie)
Implications for ATIS
It was thought that the social aspects of a site such as Cycology might be helpful for those
enduring what was described as the “shared ordeal” of travelling by bus, although it was
acknowledged that sharing negative stories could deter others from using the bus. There may
well have been an element of bias in the assumption that shared experiences about bus travel
would tend to be negative, as participants had nearly all elected to cycle to work at least some of
the time, suggesting that bus travel was not their first preference. Increasingly, bus companies
are experimenting with the use of social media, using for example Facebook and Twitter to keep
customers informed of service changes and sometimes providing a forum for feedback,
questions and discussion among users (see for example, the Facebook pages of the Devon bus
company Transdev-Yellow Buses at www.facebook.com/, accessed 15/12/10). It is estimated
that approximately 50 UK bus companies now have Facebook or Twitter pages (Austin,
2010).The sense that „social media‟ can be used to build on the sense of community among
public transport users is gaining strength among public transport companies, particularly as a
means of improving communication between providers and users (to which we return later in his
section). An example can be found in the words of Jim Allison, multimedia managing producer at
Bay Area Rapid Transit, San Fransisco:
“There‟s a natural community that exists on public transit that does not exist when people
travel by car. You get on a train and you‟re within a community. Whether you sense it or
not you‟re out there in the public with these other people, and we want to be out there in
the public as well.” (North Jersey Transportation Planning Authority and New Jersey
Institute of Technology, 2010, pp3-7).
Most people thought that information-sharing of this nature would not work for car drivers
commuting to work, because they do not consider themselves to be a community (“driving is
simply what everybody does”), and car driving might be thought of as more individualistic than
other modes. Drivers were not expected to want to share a quiet back route via a website, for
fear of this route becoming congested. Once again, this may represent a degree of anti-car bias
on the part of the Cycology participants, or perhaps a misperception, although interestingly the
majority were also motorists as well; in relation to self-categorisation theory, the „cycling ingroup‟ was more salient during interactions between researcher and participants. A more
favourable view of motorists‟ propensity to cooperate with one another might be gained from
observing the discussion forums of organisations such as the RAC (www.rac.co.uk/forum),
where motorists have shared, for example, their „favourite drives‟. An interesting recent example
is provided by drivers‟ use of an iPhone „app‟ called Tweet Park, which offers access to a large,
interactive database on parking in Israeli cities using GPS technology, and allows drivers to
„tweet‟ others in the vicinity when they vacate a parking space, “with the amorphous Tel Aviv
Implications for ATIS
parking hunting crowd turning into a community whose members help each other”
(www.haaretz.com, published 22.10.10, accessed 12/12/10).
However, Cycology participants considered lift-sharing to be the only aspect of car driving which
might benefit from informal information-sharing; here, the use of the map would be beneficial in
showing people‟s routes and possible pick-up points. It was thought that there might be potential
for developing a sense of community amongst users of a lift-sharing system, especially if this
were related to the workplace. In fact the UK-wide car-share system Liftshare provides an online
environment which can be restricted to a particular organisation. It was also noted by one
participant that motorcyclists have their own community/ies akin to those of cyclists, and might
therefore be receptive to processes of group identification through the sharing of information.
This idea is supported in the literature by the research of Esbjörnsson et al. (2004), reported in
section 2.5.
Although not usually considered as „modes of transport‟, two further potential user groups for a
system such as Cycology were thought to be wheelchair and pushchair users. The map could
provide a format for sharing detailed information about the accessibility of paths and pavements,
and could also serve a useful function in communicating problems to those responsible for path
maintenance. This leads to the final category of suggestions for the wider applicability of a
Cycology-type information system: its potential use in providing feedback to transport companies
and local authorities about everyday users‟ travel experience, communicating both the good and
the bad, and swiftly drawing attention to issues of concern at specific locations. It would remain
to be seen, however, whether opening such informal online groups to participation by what might
construed be as „officialdom‟ might change the dynamics and content of the interactions. It is
worth recalling at this point that the cycling layer of the www.bristolstreets.co.uk website, from
which the Cycology website originally grew, is used partly as a means of communicating cyclists‟
comments to the City Council.
As previously noted, public transport companies are increasingly now using social media as a
means of communicating with, and obtaining feedback from their users, either by entering into a
direct dialogue using sites such as Facebook, or through blogs and discussion forums which coexist with „official information‟ sites. A pertinent example is provided by the Bay Area Rapid
Transit (San Fransisco), which at the time of writing offered an informal blog, called Posterous,
linked to its official website:
“We're using Posterous to share tidbits of news and photos from riders. On the spectrum
of Stuff You Get From BART, this is in between our official www.bart.gov website -where you can go for trip plans, schedules, advisories etc. -- and our
Implications for ATIS
twitter.com/SFBART feed, where you can get little 140-character updates but not the
pretty pictures”.
(http://sfbart.posterous.com/ accessed 15/12/10)
Interestingly, the blog had received much less user feedback than the independent discussion
group “BART Rage”, which advertises itself thus:
“The purpose of this site is to create a community for BART riders who are willing to
share good and bad experiences. Through this site we are hoping to share with other
riders/employees and let our voices be heard to BART management that BART is
important to commuters”.
(http://bartrage.com/ accessed 15/12/10)
Regarding the potential use of a website such as Cycology for improving communication
between everyday transport users and transport providers or authorities, the comments
emerging in the interviews about the application of the technology in other areas of transport
could thus be validated through comparison with actual contemporary developments in the field
of social media and transport.
Table 11-2: Potential use of online information-sharing for other (non cycling) transport
Informal advice
Communicating with
Table 11-2 summarises the relevance of online information-sharing to different modes of
transport, as identified by Cycology participants and discussed in this section. In summary, the
route information which formed the core content of the website, and the innovative map basis,
may be a feature which can be translated most directly to walking and for specific groups such
as wheelchair users. However, the ability to post user comments on a map alongside public
transport routes and timetables might also provide a means of enhancing information and help
to create or sustain a sense of community among regular users, as well as reducing uncertainty
Implications for ATIS
for new or infrequent public transport users. The social benefits inherent in community-building
were thought to be relevant across all transport modes. Although not its primary focus (since
there are other ways of achieving this), this is a technology which was also thought to have the
potential to improve communication between ordinary travellers and providers of transport
services and authorities responsible for infrastructure maintenance.
User groups: desirable size and composition
This section addresses the following research question:
5.3 What features of a „user community‟ might be conducive to word-of-mouth informationsharing (e.g. group size, composition, types of user)?
The focus now moves from desirable social features of information systems themselves and the
potential „market‟ for user-generated information amongst users of modes other than cycling, to
considering some of those features of the user group which might contribute to maximising the
usage (and usefulness) of an interactive information system. Although a system designer may
have no actual control over who uses his or her system, greater knowledge in this area may
provide a starting point for „designing in‟ features which might encourage particular types of
contributor. This is an area which has attracted interest in fields such as information
management, where researchers are exploring the factors which motivate contributions to online
communities with a view to informing the design of technical features (e.g. Ling et al. 2005, Hall
and Graham, 2004). Such knowledge may also be useful in suggesting the types of context in
which such systems might work best (for example, an open versus a limited access website). It
may also assist those marketing and implementing such information systems (such as
workplace travel planners) by suggesting whom their most receptive „target audience‟ might be.
Firstly, some general issues about desirable group size and composition are considered, using
data arising from the participant interviews and the functioning of the Cycology website.
Consideration is then given to „ideal types‟ of user, based on a user typology constructed from a
vertical analysis of the three data sources pertaining to individual Cycology participants
(observed website activity, questionnaire and interview). To illustrate the typology, a brief
analysis of the experience and involvement in Cycology of four participants (one from each of
the four „types‟) is presented. Implications for the development of advanced traveller information
systems incorporating a „virtual community‟ dimension are discussed.
The perceived social advantages of limiting group size and composition, such as communitybuilding, trust and pro-social behaviour within Cycology, balanced against the drawback of
limited breadth of knowledge, were discussed at length in Chapter 8. It can be concluded that a
Implications for ATIS
compromise might have been achieved by increasing the overall number of participants drawn
from the five organisations; it remains a matter of speculation as to what the ideal number might
be, but it seemed feasible (and this was suggested by a number of participants) that 50 relatively
active users (contributors) might be a suitable number. „Active use‟ is difficult to define, but
experience of the Cycology project suggests that the average number of posts per participant
over the six weeks (5.7 – just under one per week), which generated an average of 4.5 posts per
day in total, was sufficient to maintain participants‟ interest in the site; it was noted in Section
8.1 that this was comparable with the level of posting activity on the Bristol Cycling Campaign
forum, which has been active since October 2006. However, the number of posts on Cycology
appeared close to the minimum level of activity necessary for it to function effectively as a forum
for information exchange and social interaction. This suggests that if user numbers were more
than doubled from 23 to 50, as was suggested, each user would need to post, on average, once
every two weeks as a minimum, although ideally more frequently. It is recognised that this
average masks the possibility that a small proportion of users might be expected to contribute
the majority of posts (as discussed in Section 10.5.2, this is borne out by the literature on online
communities, and the Cycology project itself), and that individuals‟ posts would be unevenly
spread over time. Moreover, group composition might be expected to be dynamic, as people
move out of the active user group and others join. This might be especially likely in an
environment such as a university, with a regular turnover of students.
Had a greater number of people participated in Cycology, the numbers cycling from particular
areas should have increased (thereby widening the opportunities for interaction) and the
geographical scope of the knowledge base might have been expanded by bringing in people
who cycled from other areas, allowing more questions to be answered. Participants favoured the
idea of limiting the user base to some form of defined group at a particular geographical location
– the workplace having provided an obvious example because of the commuter context of the
Cycology project. Those who worked at the university believed that such a website could be
accessible to all within it without losing its sense of community. The total number of people
cycling to the university at least occasionally has been estimated at approximately 2300 , and
this might be expected to form the core population from which active users might be drawn,
although this figure would increase if potential (as well as existing) cyclists were also attracted to
the site. In addition to the desirable number of 50 people contributing to the site at any one time,
Based on an estimated 12% of staff and 6% of students cycling to the university (2010 UWE
Travel Plan estimate). Total staff: 2,802; total students 32,359 (UWE Report and Financial
Statement, July 2010).
Implications for ATIS
it could be expected that many more might read it and learn about cycle routes without
contributing („lurking‟), extending the overall effects (for example, the static cycle map on the
UWE website received nearly 13,000 views in the two years until February 2011).
The „ideal‟ group size relates, in part, to whether or not Cycology might be considered to be an
„online community‟ at all, and whether this is a concept which might be applied to digital word-ofmouth about transport more widely. As discussed in Section 2.5 of the literature review, Kozinets
(2010) interprets Rheingold‟s (1993) assertion that “enough people” must be involved for an
online group to feel like an online community, by suggesting a minimum of 20 people. The
maximum number of people held to be possible while still allowing efficiency of communication is
between 150 and 200 (according to these criteria, Cycology was thus at the lower end). Kozinets
(2010) further argues that for an online group to become an online community it must
demonstrate more than simply the transfer of information and draw people together in
“fellowship and commonality” (p10). In this sense Cycology could be considered a „community‟
because it moved beyond the mere transmission of information, but perhaps did not go on for
long enough or involve enough people for there to be the sustained social interaction required
for participants to gain a sense of familiarity with other members – another of Kozinets‟ (2010)
criteria for defining online communities. Thus, Cycology might best be described as an
„emerging online community‟. To see whether a „mature‟ community might emerge over time,
testing is required over a longer time period, and ideally in a natural environment - making such
a system available, for example, to all employees via an organisation‟s intranet.
A further source of reflection is whether informal information-sharing in the context of advanced
traveller information really necessitates the development of an „online community‟ according to
the criteria of Rheingold (1993) and Kozinets (2010), in order to play a useful role in travel
behaviour. Hall and Graham (2004) found that norms of reciprocal helping quickly evolved in
the CipherChallenge Yahoo group, leading to effective information exchange, but without the
development of social relationships. Members were considered to be exchanging information
rather than generating knowledge, and the strength of interest in the subject (code-breaking)
was such that the group did not need strong social relationships in order to maintain its survival.
Parallels might be drawn here with travel information-sharing; arguably the „practical usefulness‟
aspect of the Cycology website could have functioned (people may still have tried out routes
suggested by others) without the added dimensions of community-building and social support,
and indeed some participants saw this instrumental function as its sole purpose. This suggests
that online information-sharing need not be restricted to specific „in-groups‟ in order to play a
useful role. However, the research for this thesis has demonstrated that the social dimension
can increase people‟s propensity to trust others‟ information, share their knowledge with others,
Implications for ATIS
and reinforce positive attitudes, intentions and behaviours with regard to cycling through
processes of referent informational influence. It may therefore increase people‟s propensity to
make practical use of the information, even if this facilitating role may be difficult to identify and
remain unacknowledged by users. Here, parallels may be drawn with the success of the
EcoTeams initiative in increasing environmentally sustainable behaviour within local
communities by bringing householders together to plan and discuss relevant activities (Staats et
al., 2004).
The conclusion to be drawn is that this is a matter of degree: whilst informal travel information,
broadcast through electronic word-of-mouth to internet users at large, may still be useful and
exert an influence on detailed aspects of travel behaviour (such as route choice), it is likely to be
less influential than if it emanated from a fellow member of a true online community. This a
matter which might be explored further through comparative research on the uses and effects of
different sources of informal, online traveller information in different contexts.
In considering the different types of individual who might together form an effective group for
information-sharing, the obvious point should be raised that different types of knowledge and
levels of experience are required within the group. If everyone were already very well informed
of local routes and highly knowledgeable about cycling issues, there would be little need to
interact with other cyclists (except for social reasons, or to collaborate in cycling campaigns –
although other means exist for doing this). Equally, if everyone were very inexperienced, there
would be little knowledge to share. It will be recalled that for this reason, one of the criteria for
the purposive selection of Cycology participants was their frequency of cycling to work, and how
long they had been doing so, in order to achieve a balance between „information-givers‟ and
„information-receivers”. Level of cycling experience did not in the end correspond directly with
people‟s propensity to engage actively with the website, and this was sometimes a surprise for
participants themselves: one person who cycled every day remarked in interview that she was
surprised to have found that, having started the project expecting to offer information more than
receive it, she had actually learned more than she felt she had contributed.
In fact, those who contributed to the website the most were not the most longstanding cyclists
(as noted in Section 8.3, two of the five most active contributors had stated cycling to work only
6 months before the project started, and two had commenced cycling between one and two
years prior to the project). These four active contributors reported at interview that they were
motivated by the social aspects of the site as well as its functional use. An interest in social ties
as well as obtaining or giving functional information was therefore a key factor in most people‟s
motivation to contribute. It should also be noted that an interest in the social dimension was not
sufficient on its own to motivate people to post messages, as those with the least cycling
Implications for ATIS
experience were also amongst the lowest contributors. A conclusion to be drawn is that an
interactive online community requires members who believe they have at least some knowledge
to share, but are also interested in building social ties. Using Kozinet‟s (2010) typology of online
communities (outlined in Section 7.9.1), a desire to engage only in practical information-sharing
about the “consumption activity” - in this case cycling - results in a “geeking community”,
whereas interest in both the activity itself and in building social ties results in a “building
community”. Hall and Graham‟s (2004) interpretation of the Cipherchallenge Yahoo group
corresponds with Kozinets “geeking community” type, and they note that although there was
cooperation and a “social infrastructure” within the group, this did not lead to genuine
collaboration and knowledge generation (something which might be expected in what Kozinets
(2010) terms a “building community”). Whilst Cycology did not, over its six week duration,
attract the level of social interaction required to become a “building community” (in fact, it was
closer to a “geeking community”, with detailed information being exchanged without the
development of strong social bonds), over half of the participants believed that they would have
been interested in both the social and functional aspects of the site had the project continued for
longer and involved more people. Together with insights from the literature on online
communities in other areas, this suggests that is desirable for an online community based
around transport to incorporate both dimensions.
A further criterion for active participation was the sense of identification with other members of
the group; it was previously noted in Section 10.2 that a minority of participants felt inhibited from
contributing because they believed everyone else to be more serious or experienced at cycling
than they were. This highlights the significant effect of social comparisons and „in-group‟
sensibilities on the dynamics within a group, and would deserve some consideration if attempts
were made to set up a „real-world‟ community of this nature to encourage the less confident
cyclist or those still only contemplating it (the same might apply in relation to communities of
other transport users).
Following discussion of the desirable size and composition of the group as a whole,
consideration is now given to the different types of individual user who might, in combination, be
most likely to maintain an active and interactive information-sharing forum, or indeed an online
11.4.1 Developing a user typology
A typology of users may prove to be of interest to those designing and implementing online
traveller information systems comprising user-generated content, by indicating the different
types of people needed to make an interactive information-sharing system function well – that is,
Implications for ATIS
attracting enough activity to sustain members‟ interest and to be considered useful. Using the
technique described in Section 7.9.1, the following typology was created by categorising
Cycology participants into two distinct dimensions: their level of observed activity on the website
(both posting and reading others‟ posts); and their level of interest in and enthusiasm for the
project (as it actually was, as well as what they believed it could become) as expressed in
questionnaires and interviews. The principle of identifying two dimensions, producing four
„types‟, was modelled on Kozinet‟s (2010) typologies of online community participation and
online community interaction, although the dimensions and types differ from those used by
Kozinets. The typology of Cycology participants is shown in Figure 11-1.
Figure 11-1: Types of Cycology participant
Types of Cycology Participant
More active
b Elaine
Marion, Laura
Julie, Rick , Mike
Sue, Adam, Rachel
Interested Lurkers
Doug, Kate, Ben
Jess, Phil, Esther
Less active
(Reported) Website Enthusiasm
By combining observed behaviour with reported „enthusiasm‟, an attempt is made to capture not
only what happened in the project, but also the potential for people‟s future engagement if they
were to use an information system such as this outside the confines of the project and without
the constraints, such as time and absence, which limited some participants‟ contribution to the
Cycology project. Participants‟ location on the vertical axis reflects an objective measure: the
Implications for ATIS
number of posts each person read and wrote. Thus, the three groups of people in the lower
quadrants (less active ) are all drawn from the bottom left of Figure 8-6, p135, which shows the
proportion of posts written and read by each individual. The three groups in the upper quadrants
(more active) are drawn from the centre and right of Figure 8-6, p135. Participants‟ location on
the horizontal axis reflects the researcher‟s interpretation of each person‟s enthusiasm for the
Cycology concept, as expressed through interviews and questionnaires, moving from
unenthusiastic on the left, to very enthusiastic on the right. An example of each type (Elaine,
Andy, Laura and Phil) has been selected for non-cross-sectional analysis in the next section.
The „active sceptics‟ and „active enthusiasts‟ comprised those who wrote the most posts, and in
all but two cases, were also those who opened the most markers. The active enthusiasts
expressed themselves to be interested in and positive about the website, whereas the two active
sceptics were more negative and said they had contributed largely through a sense of obligation.
The active enthusiasts are clustered into two groups: those who posted between 9 and 14 times,
and those who posted 6 or 7 times. The three „dissenters‟ were those who contributed little and
also expressed a lack of interest for the project (one being actively critical, the other two
unengaged). The „interested lurkers‟ were a relatively large group who wrote few posts
(between 1 and 5), but expressed enthusiasm for the project, either as it was, or what it might
become. The members of this group tended to say that they would have contributed more if they
had had more time to devote to the project. Two participants are absent from the typology as
they were not available for interview; hence their level of enthusiasm was difficult to gauge.
It is possible to speculate on which of these types might be most likely to create and sustain an
online community in the specific field of local travel; the „active enthusiasts‟ are perhaps the most
promising category, similar to Kozinet‟s (2010) “Maker” type, whom he categories as “active
builders of online communities and their related social spaces” (p34). The „active sceptics‟ could
also be expected to make a contribution initially if they felt the community to be a „good thing‟
which should be encouraged. Their longer term participation might only be anticipated, however,
if they were to move towards the „active enthusiast‟ category. „Interested lurkers‟ present a
potential reservoir of active enthusiasts, who might become more engaged under particular
circumstances (or, alternatively, might drop out and become „dissenters‟). Kozinets (2010)
describes the lurker as “a new member who is using the community to learn about the core
consumption activity or to reach out and build social relationships” (p.34). The conditions under
which they might participate more, as expressed by the participants themselves, include more
people travelling from the same area, actual availability of alternative routes, and continual
changes to the physical environment or occurrences considered „worth sharing‟. Individual
factors will also play a role; the two participants who described themselves as „unsociable‟ would
Implications for ATIS
not perhaps be expected to contribute significantly if they took little pleasure from the social
aspects of the site. One of these participants was an „active sceptic‟, but contributed out of a
sense of moral duty – exhibiting pro-social behaviour - rather than through enjoyment. This
suggests that pro-social behaviour may not be sustained if there is no sense of reciprocal benefit
from participating.
It is also useful to consider the attributes which people said they wanted from this or a similar
site. Some people described it as “not a social site”, remarking that they would only use it in the
longer term if they had a specific question, such as looking for a new route, and would not
engage in “chit chat”. For them, the purpose of the site was to share practical information, not
social information conveying mutual support, encouragement for cycling and so on. They
regarded it as a resource which they might use if and when they felt the need, rather than
regarding it as an online community as such. These were, in the main, participants who were
long-standing cyclists, with relatively fixed cycling behaviours, and therefore considered
themselves to receive little benefit from community-building and supportive aspects of the
website – although they saw value in these features in principle as a means of encouraging
novice cyclists. As discussed in the previous chapter, those who appeared to gain most from the
social and supportive aspects of the website tended to be those who had more recently taken up
cycling to work. These also tended to be the people who were looking for stronger social ties, or
at least more frequent social interaction, on the website (although some long-standing cyclists
also fell into this category). Arguably these are the types of people who might be expected to
make a greater contribution to sustaining an online community, and are all to be found within the
categories of „active enthusiast‟ or „interested lurker‟.
11.4.2 Four individual cases (of user)
In order to gain a deeper understanding of the beliefs, attitudes and behaviours of the Cycology
participants, a holistic, or vertical analysis of all data sources pertaining to each individual was
carried out. Mason (2002) describes this form of analysis as “a practice guided by a search both
for the particular in context rather than the common and consistent, and the holistic rather than
the cross-sectional” (p.165). Freudendal-Pedersen et al. (2010) describe it as a means of
bringing out individual voices from within the data. This method was described in Section 7.9.2.
The cases of four Cycology participants are summarised in Boxes 11.1 to 11.4, each illustrating
one of the four „types‟ within the typology. The aim is to highlight, within the context of the
„whole person‟, some of the attributes which may be conducive to active, online informationsharing.
Implications for ATIS
The factors we have discussed in this part of the chapter, relating to the desirable size, nature
and composition of a user group which might interact to cooperate and share information are
summarised below:
Number of active contributors: approximately 50;
User base: a defined group (e.g. a large employer or co-located smaller organisations).
User attributes:
People with different levels of knowledge and experience of the relevant transport mode
– including those with none („enough‟ of each so that no category feels excluded);
People interested in social interaction as well as functional use.
User types: (in order of importance):
active enthusiasts, interested lurkers, active sceptics. The active enthusiast types
corresponds with Kozinet‟s (2010) Maker – the active builders of online communities,
whilst interested lurkers correspond with his Lurker, who observes, learns and feeds into
the community.
An interesting next step, which remains outside the scope of thesis, would be to develop a
strategy for recruiting the different types of people, as well as reflecting on the sort of system
design features which might encourage the different types to use it.
Implications for ATIS
Box 11.1: “Active Sceptic”
Elaine (30-39, cycles to work every day, began 2.5 years previously)
Elaine said in the interview that she was only interested in the practical usefulness of the site,
not "chit chat". She did not see Cycology as a place for socialising - it was simply a place to
share practical information. She describes herself as a private person who does not like to
"share bits of herself". Hence she does not take part in internet discussion groups and does not
wish to socialise on a website. She believes that some people like to feel part of a group or
community, but she does not because she considers herself unsociable. Similarly, she cycles
for her own reasons and does not need to feel part of a collective. However, she said she was
quite happy to ask and answer specific questions, both face-to-face and electronically. She had
used a considerable amount of word-of-mouth information when she first started cycling, and
likes to encourage others to cycle by sharing her knowledge. Although she was very clear about
making her own decisions about cycling, she was not entirely sure that the cycling culture and
peer group pressure at her place of work had not influenced her in some way.
Ironically (as she thought she was not very involved and was only participating out of a sense
of duty), she was one of the more interactive participants, posting 11 times, responding to other
comments 4 times and also asking questions. Her contributions included a high degree of
subjective opinion and advice, and were written in an informal style. Although she said that she
did not like contributing to “chit chat” or writing "isn't it nice to see the birds", but did write a
descriptive post of this nature (the cat which likes to be tickled) which others remembered and
liked. She logged in on 7 days, opening a total of 44 markers. This was a case where there was
a discrepancy between the observed level of posting activity and varied content (she was one
of the most active contributors), and her own perception that she had not been especially
engaged with the project.
Elaine slightly contradicted herself by saying that she was mainly contributing just for the sake
of the research project, but then said that she would probably also have contributed if it were an
open project within her place of employment (i.e. not a research project). She also thought that
she would contribute to a public website such as www.bristolstreets.co.uk , all out of a sense of
civic duty. She might therefore still be an active contributor to an interactive system outside the
research project context even if she were not interested in the social dimension, perhaps due to
her „pro-social‟ values. However, without genuine interest, her involvement might not be
Implications for ATIS
Box 11.2: “Dissenter”
Andy (30-39, cycles to work infrequently, began 2.5 years previously)
Andy was markedly more negative than any of the other participants about the website - and,
since the majority was generally positive, could be described as an “outlier”. There appeared
to be for two main reasons for this : 1) he did not like the way the website worked technically,
as he thought the bubbles on the map format and unstructured digest did not stimulate
interaction; 2) he seemed to feel alienated from the other participants, which contrasted with
most other people in the project. The latter point seemed to emerge from his experience of
the project briefing, where he had formed the impression that many of the others were
“hardcore cyclists carrying their helmets and wearing special shoes”. He stressed the point a
number of times that he was not a very enthusiastic cyclist, and felt that the other participants
were – hence he was bored by their comments rather than feeling any empathy with them. He
had no wish to exchange banter or bond with them, and was only interested in obtaining
functional information, not to engage in sustained interactions or community-building. He said
that this would have been different if he had known the other participants, but as it was, he
“did not care” what they thought.
One possible factor for this disengagement many be that Andy, during a chance encounter
with the researcher after the project, admitted that he had seldom cycled during the project,
having realised how easy it is to park a car at his place of work in the summer. He had
previously been very enthusiastic about cycling, which suggested that a degree of cognitive
dissonance may have occurred (changing attitudes to match behaviour).
Andy also thought he felt negative about the website because he is very used to a discussion
forum format, which he believes to be much more conducive to interaction. He did like the
Cycology map, but thought it would be better if the forum were the focus, with links to a map.
Because it was unclear to him whether people were responding to threads, and because he
was not engaged, he did not realise that there had been three responses to his question
about a difficult junction. He actually found this quite interesting when he read it in the
interview, but blamed the format for the fact that he had missed them. However, although
Andy was generally negative about the project, he was positive about the overall concept,
saying that he liked the map and that a university-wide initiative was a great idea which might
encourage more people to cycle. He thought that informal information-sharing would also
work well for bus users, although he personally would only use if for instrumental information.
Implications for ATIS
Box 11.3: “Active Enthusiast”
Laura (20-29, frequently cycles to work, began 18 months previously)
Like some of the other participants from a smaller organisation, Laura worked out the
identity of others in the project. The much smaller size of her organisation made this more
possible than it was for participants from the larger organisations. Laura emerged, from the
interview and questionnaire, as possibly the most enthusiastic of the participants. She
concentrated on the more community-building, sharing-experience type of posting, and was
able to remember more specific examples than anyone else. She also liked the practical
advice and warnings, but was most enthusiastic about comments which created a positive
impression of a route and pointed out features which helped her to enjoy the experience,
such as the black cat which liked to be tickled. She opened 84 markers over 7 days and
posted 11 times - a combination of routes, warnings, questions and a "gripe"- making her
one of the most active contributors. Her tone was very informal and conversational.
Interestingly, although she mentioned seeing someone cycling along whom she recognised
from work, and also worked out that she was in the project, she did not mention the fact that
she and another participant had met up socially after meeting one another for the first time
during the project briefing (this person described Laura as a friend during her own interview).
Perhaps this was because Laura had not really associated the project briefing with the
project itself, or perhaps because her friend was not very visible on the website.
Laura liked the small size of the project, feeling that the group was small enough for her to
start recognising names (although she had not really got any sense of them as individuals).
She thought that there was enough interaction and a sufficient number of people involved.
One reason might be that she does not talk to other cyclists at her workplace very often, so
Cycology served as a genuine "social space" for her. She was the only person who used a
pseudonym on the site, and she seemed to like this level of anonymity. Although she did not
try a new route, she said that the project has changed her experience of cycling, making her
more observant in looking out for pleasant features, or hazards, to pass on to others. She
was enthusiastic enough in this respect to take photos of both hazards and attractive views,
but was unable to upload them to the site. She had tried to make looking at the site/email as
part of her daily routine, and found it a pleasant distraction when she needed a break from
her work. She said that the project had also made it more likely that she would join a cycling
discussion forum. She described Cycology as a "virtual community of people who cycle".
Implications for ATIS
Box 11.4: “Interested Lurker”
Phil (30-39, cycles to work frequently, began 6 years previously)
Phil was a participant who, from observation of website activity, appeared to have little
interest, but showed through the interview and questionnaire that he was actually very
enthusiastic, saying that he had regularly read the email digest to follow developments on the
website. He made two contributions: once to reply to a thread about why the gates on the
rugby ground route are sometimes locked (providing an explanation); and once to mark his
route to work from home, commenting on lighting, traffic and steepness. He logged in just
once, opening 11 markers. In both the interview and questionnaire he said he had participated
less than he intended and this was due to time constraints.
After the project, Phil publicised www.bristolstreets.co.uk within his company later in the
summer, when a new cohort of several hundred people started working in his building, in
order to encourage them to cycle, and this was probably one of the largest practical outcomes
of the project in terms of word-of-mouth information diffusion.
Despite being a long-standing commuter cyclist, Phil was a strong proponent of the
"community dimension" of the website, and raised this early in the interview without any
prompting. He liked the way that the project gave him a sense that others were cycling his
route, as he travels to work early in the morning and sees few people en route. For him, the
project provided a pleasant feeling of shared experience. He thought a practical outcome
might be that people would be more likely to raise infrastructure problems with the route with
the Council, because they would be more aware that others would be experiencing the same
problems. He gave an example of feeling more confident to raise an issue with a Council
official because he knew that many others were affected.
He is enthusiastic about cycling as means of transport and likes to encourage others to do so
- he gave examples where he believed he had played a key role in two people's decisions to
start cycling to work (his wife, who now cycles to her place of work, and a colleague in his own
company). He tried out one of the Cycology routes to UWE with his wife; hence, electronic
word-of-mouth had been followed by face-to-face word-of-mouth influence.
Implications for ATIS
11.5 Chapter Summary
This chapter has sought to answer research question 5 by considering some of the ways in
which social aspects of the Cycology website might be translated to other web-based traveller
information systems, both for cycling and for other modes of transport. Additionally, it has
considered features of the user group and types of individual within it which might make
information-sharing as effective as possible. This includes creating an environment where there
is enough interaction, for enough knowledge to be shared to be of direct practical benefit to
members, as well as creating a sense of community where processes of social influence are
Recommendations for the incorporation of „social design features‟ into other forms of advanced
traveller information system highlighted in particular the benefits of combining user comments
with routes generated by automated planners. User-generated route information was considered
to be especially relevant for cycling and walking, but general user comments on transport
services were also considered to be potentially useful for public transport. Comments could be
posted on the map alongside public transport routes, and it was thought that this might be useful
not only for other travellers (especially those who were new to, or considering a mode of
transport), but also in terms of providing feedback to transport operators. Similarly, user
comments posted to the map by cyclists and pedestrians (as well as wheelchair and pushchair
users) could communicate both good and bad features of the route to those responsible for
maintaining them. The community-building aspects of this type of system were thought to be
beneficial in terms of motivating and sustaining cycling, walking, lift-sharing and the use of public
transport. Community-building was thought most likely to occur within distinct groups where
members shared a sense of identity. In order to sustain enough interaction to create or maintain
such a sense of community, approximately 50 contributing members were thought to be
required. Four „types‟ of website user were identified in the Cycology project, and it was
concluded that „active enthusiasts‟ were particularly important in sustaining interaction, although
„active sceptics‟ might also be expected to contribute for pro-social reasons, and „interested
lurkers‟ provided a reservoir of potentially more active users.
Chapter 12 : Conclusions
In this final chapter, the main conclusions of the thesis are drawn together in two areas: theory
development; and practical applications to advanced traveller information systems. Wider
discussion points emerging from the research, including relevance for policy and measures to
encourage travel behaviour change, as well as recommended areas for further research, are
then considered.
12.1 Theory development
This thesis has applied constructs and theories from social psychology to improve understanding
of the use and effects of traveller information from a perspective which has hitherto been
neglected: the role of informal information; its transmission through social interaction; and
related influences on everyday travel behaviour. Whereas „conventional‟ understandings of
travel information focus on „facts‟ about, for example, times, routes and costs, provided by
official sources to help the individual make utility-maximising travel choices, this analysis has
conceptualised travel information as something broader in which „facts‟ are overlaid with
subjective opinions, emotions and normative messages as they are communicated between
people. The addition of a „social layer‟ to the travel information means that social processes are
also in operation alongside well-documented processes of individual, instrumental reasoning. In
the case-study, the interactive nature of the website, together with the limited size and
composition of the user group created an environment for processes of group identification,
trust, pro-social behaviour and social influence. This suggests that some aspects of the
conventional, individualist paradigm within which traveller information has been conceptualised,
should be questioned. Moreover, this thesis has shown - through participant accounts in both
phases of the research - that the role of information communicated through word-of-mouth does
not begin when an active decision-making process has been initiated, but also plays a more
subtle, background role in the formulation of beliefs and attitudes, which in turn shape the
context in which active travel choices may later be made. Furthermore, where information is
obtained from others within a reference group, it plays not only a straightforward informational
role, but also a normative one by conveying norms of preferred behaviour within that group;
information of this nature is perceived as more valid than information from other sources. The
process of providing information to others is also affected by social factors such as reciprocal
At the same time, it was clear from both stages of the research that informal items of information
and word-of-mouth information diffusion were just two small elements within the complex
decision processes affecting everyday travel behaviour, and that these processes are essentially
(but not exclusively) „individual‟ and „rational‟ – for example, there was no evidence of people
„blindly‟ following the advice of others without deliberation. Factors such as instrumental
constraints, personality-related factors and conventional notions of individual, utility-maximising
travel behaviour were also apparent in the findings. In fact, social identity theory has been
criticised for assuming too sharp a distinction between personal and social identities (Wetherell,
1996). Therefore, the new knowledge generated through this research about social influence
through word-of-mouth information-sharing seeks to complement, but not replace, existing
models of the role of information within the travel decision process.
Figure 12-1 shows how the social factors discussed in this thesis, informed by types of theory
not previously used in this field, and validated in an applied context, can be integrated into an
information processing model (e.g. Bettman, 1979, cited in Gabbott and Hogg, 1994; Jackson
and Polak, 1997; Sirakaya and Woodside, 2004; van der Horst, 2006), thus adding to existing
theory. This model takes one part - the Section to the lower left - of the process model which
was developed at the end of Phase 1 (Figure 6-1, p95), and elaborates it on the basis of the
more detailed social-psychological findings in Phase 2. It also represents a form of validation of
the more general Phase 1 findings within a specific „real world‟ context – that of a traveller
information system used by a group of cyclists.
The model depicted in Figure 12-1 is an elaboration of the second stage of an information
processing model: information search. It will be recalled from Section 2.1.2 of the literature
review that this stage follows problem definition, and precedes evaluation and selection;
implementation; and post-hoc evaluation. As previously noted, the research for this thesis found
that word-of-mouth information also plays a role in travel decision processes prior to problem
recognition, when social interaction about travel can reinforce social norms of behaviour and
contribute to the formulation of an individual‟s beliefs and attitudes through a process of “passive
information-catching” (Um and Crompton, 1990). This is important because it may influence the
types and sources of information (both formal and informal) which are later sought when an
active decision process is initiated, i.e. at the stage of problem recognition.
Figure 12-1 : Processing of informal and formal travel information
Internal search
External search
e.g. online
e.g. Fig. 12.2
e.g. routeplanner
Social factors:
Group identification
(+group attributes)
Social judgement
Reciprocal helping
Individual cognitive
Social cognitive processing
Referent informational influence
Evaluation and
(e.g. route choice)
of behaviour
In existing models, information search, is, in turn, divided into internal and external information
search (Jackson and Polak, 1997; Avineri and Prashker, 2006). In the former, an individual
draws on memory of past experience, and only if this provides insufficient information will he or
she execute an external information search. Here, Figure 12-1 extends existing models by
dividing external information into: formal types of information obtained directly from formal
sources (e.g. timetables and journey planners); and both formal and informal types of
information obtained through word-of-mouth. An example of an online information source within
each of these categories is shown in the three circles. The circle to the left denotes existing
„bottom-up‟ information sources such as the travel review website www.tripadviser.com or the
cyclists‟ social network http://morvelo.cc/. The circle to the right denotes existing „top-down‟,
electronic information sources such www.transportdirect.info. Of greatest relevance to this thesis
is the form of combined system denoted in the circle in the centre of the figure. This is the type
of nascent, online system which incorporates both top-down (formal) and bottom-up (informal)
information, exemplified by the webpage mock-ups in Figure 12-2, which will be described in the
next section.
Figure 12-1 shows some of the social-psychological factors which may be involved when both
formal and informal types of information are communicated through word-of-mouth. Particularly
in the case-study, it was found that the more important issue was not whether information was
formal or informal in content (in fact, the two were often mixed), but the fact that it was delivered
through word-of-mouth rather than from „official‟ (formal) sources. Factors such as group
identification and trust may increase the influence of information conveyed through word-ofmouth if it is provided by a person perceived as similar – i.e. if it comes from a member of a
reference group. In this case, the social transfer of information may provide a channel of
“referent informational influence” (Turner at al., 1987). Self categorisation theory suggests that
such information is subject not to individual cognitive processing, as conventional
conceptualisations of information-use suggest, but to social cognitive processing (Turner et al.,
1987). The stages of this process are denoted by the thicker arrows in Figure 12-1, and
represent the main area of findings in this thesis.
However, the research did not show that information obtained through word-of-mouth is
necessarily always affected by social factors; in some circumstances, it may be subject to
individual processing in much the same way as information from formal sources. These
processes are represented by the thinner arrows to the right of the model. This might be the
case if, for example, the information-seeker feels no social connection with the information-giver,
or is not receptive to the social aspects of the information, as was the case with some of the
routine cyclists in the case-study who stated that they were interested only in instrumental
aspects of the website. Examples provided by participants in the exploratory research included
fleeting conversations with fellow travellers when instrumental information was sought or offered
during a trip. Trip context was also important in this respect: social factors played a stronger role
in word-of-mouth information-use for those travel decisions with a higher cost – in terms of
money, time or effort – than it did for „low risk‟ decisions, when the costs of acting on information
which turned out to be unreliable were low.
The individual use of formal travel information from official sources, on the right of the figure,
was not explored in any detail in this research (this area having been widely studied already),
but it frequently appeared in participant accounts by way of contrast with word-of-mouth
information. It was reported by many participants that information from a known and trusted
(personal) source was given greater weight than information from an official source; for example
transport companies and local authorities might be thought to “have an agenda”, and/or lack the
detailed and up-to-date knowledge which a local transport-user could provide. However, overall,
informal and formal sources and types of information were thought to serve different purposes
and to be complementary (hence the expressed wish to be able to combine automated route
planners or timetables with user-generated comments).
Within an information processing model, all these sources and types of external information then
feed into the next steps in the process: an evaluation of alternatives; the selection and
implementation of a decision (e.g. a travel behaviour); and post-hoc evaluation. Within the casestudy, this was exemplified by the selection, trying out, and assessment of new cycle routes.
However, this thesis has also explored the broader role of word-of-mouth information, showing
that, whilst it does not always lead directly to the active selection of a behaviour, it may
nonetheless exert a more subtle influence, for example by providing confirmation of existing
travel choices, and these factors may indirectly influence behaviour. Hence, the sense of mutual
support and empathy experienced by some cyclists in the Cycology project was reported to have
reinforced pro-cycling attitudes and intentions, which in turn encouraged them to maintain their
cycling activity.
12.2 Advanced traveller information systems
This section discusses two higher-level findings which arose from the research, which may have
relevance for future developments in advanced traveller information – particularly web-based
systems offering real-time travel information and journey planning. The two principles discussed
below are: that the incorporation of bottom-up, informal information can enhance the perceived
quality and reliability of information systems; and that such mechanisms can provide transport
providers with useful feedback from users. An example is then provided of a way in which
practical recommendations reported in Chapter 11 might be implemented through the adaptation
of a cycle route planner within a particular organisation.
The first broad principle - outlined in the previous section - is that user-generated information
often inspires greater trust than „official‟ information, largely because it is based on the
experience of „real people‟ rather than transport service providers or governmental bodies. This
involves both calculus trust, emanating from the intrinsic quality of the information (e.g. detailed
and up-to-date), and relational trust, associated with the relationship between information-giver
and receiver (Rousseau et al., 1998); however, the former trust mechanism is primary. Thus, the
users of the Cycology system considered the information to be more reliable (although less
comprehensive) than the information they might acquire from a standard cycling map or routeplanner. The case-study also provided evidence of people enacting a new travel behaviour (in
the form of using new cycle route) on the basis of the information provided. Although not tested
in this research, this raises the possibility that the incorporation of user-generated information
may increase the perceived reliability of traveller information systems more generally, which
may, in turn, enhance the effects of such systems on travel behaviour. This thesis has
suggested that this is most likely to occur if information systems can be „localised‟, allowing
processes of group identification to occur amongst users – for example by linking journey
planning services or real-time travel updates (incorporating user comments) to the websites of
individual organisations („white-labelling‟), or allowing restricted-access discussion groups to
form alongside top-down, open-access information services. Indeed, this is consistent with
recommendations in the Local Transport White Paper (DfT, 2011) that the Transport Direct cycle
journey planner be used within local contexts such as schools, social groups, railway stations,
hospitals and universities; moreover, the White Paper recommends that local authorities
promote it as a workplace travel planning tool.
The second principle is that information contributed by travellers is useful not only to other
travellers, but also to transport providers (and transport planners), as it offers a means of
updating them on day-to-day issues such as occurrences on the transport network, infrastructure
problems (e.g. roads, cycle paths and footpaths), or errors in journey-planning information. The
findings suggested that there is a demand from users for systems which allow prompt and
efficient communication - aided by an online map to locate problems – not only between
transport users themselves, but also between users and providers. In practice, this might mean
encouraging representatives of local transport service providers to access the „localised‟ traveller
information systems which we have described above.
In addition to these general principles, more detailed, practical recommendations for „social
design features‟ also emerged from the case-study, which may be applicable to a wider range of
web-based information systems – particularly those where a map forms a significant part of the
interface (for example, map-based journey planners, or real-time, location-based information,
such as maps showing the impending arrivals of buses at specific stops). Suggestions for the
development of design features, particularly relevant to route planners for walking, cycling and
car-sharing, were described in Chapter 11. User comments could also be helpful where public
transport routes, or live incidents on the network, are presented on a map. Other
recommendations included the incorporation of more social networking-type features, such as
user profiles, into the interface.
Figure 12-2 provides a visualisation, using four simplified, webpage mock-ups, of how a cycling
route planner might be adapted to incorporate features of this type within a particular
organisation. Transport Direct‟s cycle route planner is currently „white-labelled‟ on the University
of the West of England‟s (UWE) website, but the mock-ups show how the system might be
developed to allow users within the university to add their comments to sections of route
generated by the route planner, read other users‟ comments, and discuss related issues. The
first page mock-up is thus an amended version of the existing UWE webpage which connects to
the Transport Direct route planner, but with the addition of references and hyperlinks to user
comments and discussion. Page 2 shows part of a route to the university generated by the route
planner, with written instructions appearing to the left of the map. The new feature is that a user
might then click on the link under a section description to open markers created by other UWE
users near that section of the route, or add their own markers. Page 3 shows a comment created
by a UWE cyclist near the first section of route described on the left side-bar (direction 2.6). The
marker also includes a link to the contributor‟s user profile – a social-networking feature. Having
opened this marker, a reader may then click to read responses to this comment (message
thread), which appear on the left side-bar (page 4). The reader can then add to the thread, or
start or open other discussion threads. If the user comments were accessible only to users with
a UWE log-in (whilst leaving the standard route planning service accessible to all, so as not to
exclude, for example, visitors to the university), the social benefits arising from a „socially
proximate‟ group of users – described in detail in the previous two chapters, might be expected
to ensue.
Figure 12-2: Mock up of a cycle route planner incorporating user comments
Page 1
Page 2
Page 3
Page 4
In conclusion, if the range, quality, and perceived reliability of traveller information can be
improved through the incorporation of user-generated content - and this research suggests that
it can - it may be a matter which warrants consideration for a wider range of advanced traveller
information systems, as the technological opportunities for doing so continue to grow.
12.3 Discussion: relevance to behavioural change policies and measures
Following the presentation of theoretical and applied conclusions, in this final section we return
to the wider policy context within which the research was originally framed – as set out in the
introduction to the thesis. Can any conclusions be drawn from the research with regard to the
development of user-generated traveller information as a tool to encourage sustainable travel
behaviour and help tackle traffic congestion? In discussing this matter, we also consider areas
of further research which are required in order to achieve better understanding in this area.
The research demonstrated that although word-of-mouth traveller information plays a wider
social and psychological role than a direct influence on the selection of travel choices, it is
indeed one factor which can contribute to the selection of a travel behaviour, particularly with
regard to details such as route choice. Examples of trip details which were influenced by wordof-mouth were reported by participants in the first phase of the research, and this was directly
observed in the case-study, where over half of the participants used a cycle route suggested by
another participant. However, as we have acknowledged, word-of-mouth influences are clearly
just one amongst a myriad of other factors which affect everyday travel behaviour. Furthermore,
where influence does occur, it is not always easy to distinguish straightforward informational
influence (when information meets a practical need and is used regardless of its source) from
normative influence (where social factors also affect the use of the information). Further
research, perhaps of a quantitative nature, would be required to test whether or not there is a
direct and significant relationship between the specific social factors discussed in this thesis, and
people‟s propensity to follow the travel advice of others. Moreover, it might be conjectured that
cyclists – the main focus of this study - are particularly susceptible to notions of in-group
identification, so that further work is required to understand the relevance of this and other
social-psychological theories amongst users of other transport modes, in different social
contexts beyond the limited socio-economic scope of this study, and with larger samples.
Caveats aside, it is possible to reflect on the relevance of some of the findings to policies and
measures designed to encourage sustainable travel behaviour. The exploratory research phase
indicated that significant behavioural shifts (such as change of transport mode) are unlikely to
result directly or exclusively from word-of-mouth interactions, and the Cycology project was
unable to challenge this finding, as nearly all participants already cycled to work. However, even
those individuals who were less enthusiastic about their own involvement in the project reported
that they believed an initiative such as Cycology, if extended for general use within an
organisation, could be one factor in encouraging people to take up cycling to work. Further
research would be desirable, especially amongst people considering, but not yet cycling (or
indeed amongst those considering the use of other sustainable transport modes), to explore this
further. Whilst the case-study did not generate clear findings about the role of word-of-mouth in
encouraging people to start to cycle, it did, however, show that it could help sustain those who
had started relatively recently. The process of sharing information could perform a communitybuilding role whereby positive views of cycling as a commuter mode were reinforced, alongside
the more obvious functional role of diffusing practical travel information. Both roles were thought
to offer particular encouragement to those who were new to cycling or new to a particular
workplace, when identification with other cyclists appeared to be especially salient. This is
interesting if one recognises the importance of encouraging people not only to start cycling, but
also to maintain it once they have started. A recent study for the DfT: Cycling, Safety and
Sharing the Road (Christmas et al., 2010) notes that the reasons why a person retains a habit
are not always the same as their reasons for originally acquiring it:
“This serves as a reminder that, even when we have found the real reasons why people
cycle, promoting cycling and sustaining cycling are not the same thing”. (p.17)
Thus, the role of informal information-sharing in helping people to maintain a sustainable travel
behaviour is worthy of consideration in the overall context of behavioural change.
The importance of group identification in the Phase 2 findings also gives rise to the question of
whether it is possible to stimulate „sustainable transport identities‟ as a means of encouraging
sustainable travel behaviour within defined communities such as the workplace, and whether
this principle could be developed within the context of travel plans. Such processes may, indeed,
already be at work through established initiatives such as communal “bike breakfasts” for
cyclists during the annual “Bike to Work Week” in the UK (www.bikeweek.org). One challenge
would be to stimulate such identities in a way which draws people in, rather than creating
exclusivity; for example, a reference group of people who walk, cycle or use the bus occasionally
might be more conducive to travel behaviour-change overall, than would the existence of
stereotypical „hardcore cyclist‟ in-groups, which might risk alienating potential new cyclists .
Finally, this discussion might be placed within the context of contemporary policy interest in both
“nudge theory” (Thaler and Sunstein, 2009, as discussed in the Local Transport White Paper,
DfT, 2011), and in understanding how policy „nudges‟ towards socially desirable behaviours can
permeate through social networks - or alternatively, why some nudges have no effect at all
(Ormerod, 2010). An information-sharing forum such as Cycology may act as a nudge by both
increasing choice through the provision of instrumental travel information, but also by
communicating normative information about the transport choices which „similar others‟ are
making. It also provides a form of social network through which such nudges can quickly be
diffused. Information flows between members of the Cycology group were briefly touched upon
in Chapter 8, and there is considerable scope for research into the diffusion of travel information
through social networks, and the effects of such information on travel behaviour amongst
network members. Greater understanding of network effects may, in turn, contribute to more
effective policy-making. As Ormerod (2010) argues:
“when it comes to contemporary challenge – climate change for example – it seems
clear that we will often need to induce dramatic mass behaviour change. We are unlikely
to do so using simple incentive based approaches and need to get better at harnessing
the power of networks.” (p.37)
Drawing on a range of social-psychology theories, and using an exploratory, qualitative
approach, this thesis has begun to shed light on the role of „spreading the word‟ about everyday
travel behaviour, especially through the medium of internet technologies, but in this multi-faceted
and fast-changing area, there remain new avenues to explore.
Accent Marketing and Research (2002). Positioning Products in the Information Marketplace –
Phase 2 Report. Report to the Transport Direct Programme, Department for Transport. London.
Adar, E. and Huberman, B.A. (2000). Free riding on Gnutella. First Monday [online], vol. 5,
no.10. [Accessed 25/06/10].
AEA Technology (2007). Transport Direct Evaluation: Final Report for the Department for
Transport. Didcot, Oxfordshire.
Ajzen, I. (1991). The Theory of Planned Behaviour. Organisational Behaviour and Human
Decision Processes. 50, (2), pp.179-211.
Ajzen, I. and Fishbein, M. (1977). Attitude-behavior relations: A theoretical analysis and review
of empirical research. Psychological Bulletin. 84, pp.888-918.
Allport, G.W. (1962). The general and the unique in psychological science. Journal of
Personality. 30, pp.405-22.
Ampt, E. (2003).Voluntary Household Travel Behavioural Change – Theory and Practice. 10
International Conference on Travel Behaviour Research, 2003, Lucerne.
Anable, J. and Gatersleben, B. (2005). All work and no play? The role of instrumental and
affective factors in work and leisure journeys by different travel modes. Transportation Research
Part A. 39, pp.163-181.
Anable, J., Lane, B., and Kelay, T., (July 2006). An Evidence Base Review of public attitudes to
climate change and transport: Final report to the Department for Transport.
Arsal,I., Woosnam, K.M., Baldwin, E.D. and Backman, S.J. (2010). Residents as Travel
Destination Information Providers: An On-line Community Perspective. Journal of Travel
Research, 49 (4), pp.400-413.
Ashworth, P. (2008). Conceptual Foundations of Qualitative Psychology. In Smith, J. A., ed.
(2008), Qualitative Psychology: A Practical Guide to Research Methods. 2
Ed. London: Sage,
pp. 4-25.
Austin, J. (2010). Use of Social Networking to promote public transport and sustainable travel.
Local Transport Today, ITS Supplement. Autumn 2010.
Avineri, E. and Prashker, J. (2006). The impact of travel time information on travelers‟ learning
under uncertainty. Transportation. 33, pp.393-408.
Avineri, E. and Goodwin, P. (Eds.) (2010). Individual Behaviour Change: Evidence in Transport
and Public Health. Project Report for the Department for Transport, London.
Axhausen, K.W. (2003) Social networks and travel: Some hypotheses, Arbeitsbericht Verkehrsund Raumplanung, 197, Institut für Verkehrsplanung und Transportsysteme, ETH Zürich.
Axup, J. , Villier, S. and Bidwell, N. (2005). Usability of a Mobile, Group Communication
Prototype While Rendezvousing. CTS‟05 International Symposium on Collaborative
Technologies and Systems- Special Session on Mobile Collaborative Work, 15-20 May, St
Louis, MO, USA.
Axup, J. and Viller S. (2006) Technical Report No.463, Lo-fi Matchmaking: a Study of Social
Pairing for Backpackers, School of Information Technology and Electrical Engineering,
University of Queensland.
Bagozzi, R, Gürnao-Canli, Z. and Priester, J. (2002). The Social Psychology of Consumer
Behaviour. Buckingham: Open University Press.
Bagozzi, R. and Lee, K-H. (2002). Multiple routes for social influence: the role of compliance,
internalisation and social identity. Social Psychology Quarterly.65 (3), pp.226-247.
Bamberg, S., Rölle, D. and Weber, C. (2003). Does habitual car use not lead to more resistance
to change of travel mode? Transportation. 30, pp. 97-108.
Bamberg, S. and Schmidt, P, (2003). Incentives, morality or habit: predicting students‟ car use
for university routes with the models of Ajzen, Schwartz and Triandis, Environment and
Behaviour. 35 (2), pp. 264-285.
Bandura, A. (1977). Social Learning Theory. New Jersey: Prentice-Hall Inc.
Banister, P., Burman, E., Parker, I., Taylor, M. and Tindall, C. (1994). Qualitative Methods in
Psychology. Buckingham: Open University Press.
Beale, J.R. and Bonsall, P.W. (2007). Marketing in the bus industry: A psychological
interpretation of some attitudinal and behavioural outcomes. Transportation Research Part F. 10
(4), pp.271-287.
Ben-Akiva, M.J., Bowman, J., and Gopinath, D. (1994). Travel demand model system for the
information era. Cambridge, MA, Massachusetts Institute of Technology.
Bettman, J.R (1979). An Information Processing Theory of Consumer Choice. Massachusetts:
Addison-Wesley Publishing Company.
Bettman, J.R., Luce, M.F. and Payne, J.W. (1998). Constructive Consumer Choice Processes.
The Journal of Consumer Research. 25 (3), pp. 187-217.
Bieger, T., and Laesser, C. (2004). Information Sources for Travel Decisions: Towards a Source
Process Model. Journal of Travel Research. 42, pp. 357-371.
Bierhoff, H. W. (2001). Prosocial Behaviour. In Hewstone, M., Stroebe, W., eds. (2001).
Introduction to Social Psychology, Oxford: Blackwell, pp.285-314.
Bierhoff, H.W. (2002). Prosocial Behaviour. Philadelphia: Psychology Press.
Blaikie, N. (2000). Designing Social Research. Cambridge: Polity Press.
Boarnet, M.G., Anderson, C.L., Day, K., McMillan, T., and Alfonzo, M. (2005). Evaluation of the
California Safe Routes to School Legislation. Urban form changes and children‟s active
transportation to school. American Journal of Preventative Medicine. 28 (2), pp.134-140.
Boden, D.. and Molotch, H. (1994). The Compulsion to Proximity. In Friedland, R. and Boden,
D., eds.(1994), NowHere: Space, Time and Modernity. Berkeley: University of California Press.
Bohner, G. (2001). Attitudes. In Hewstone, M. and Stroebe, W., eds.(2001), Introduction to
Social Psychology: a European Perspective. Oxford: Blackwell. pp.239-282.
Bonsall, P. (2009). Particular Problems to be Overcome when Seeking Data on Sustainable
Travel Behaviour. Annual Conference of the Transportation Research Board, January 2009,
Washington D.C.
Braun, V. and Clarke, V. (2006). Using thematic analysis in psychology. Qualitative Research in
Psychology. 3, pp.77-101.
Cairns, S., Sloman, L., Newson, C., Anable, J., Kirkbride, A. and Goodwin, P. (2004). Smarter
Choices - Changing the Way we Travel. London: Department for Transport.
Carrasco, J. A. and Miller, E.J. (2006) Exploring the propensity to perform social activities: a
social network approach. Transportation. 33, pp.463-480.
Casaló, L.V., Flavián, C., and Guinalíu. M. (2011). Understanding the intention to follow the
advice obtained in an online travel community. Computers in Human Behavior. 27 (2), pp. 622633.
Chabot, A. (2007). Travel 2.0: future of tourism? An investigation into the web 2.0 and its impact
on the travel and tourism industry. Masters dissertation, Bournemouth University and
Fachhochschule Heilbronn.
Charmaz, K. (2005). Grounded Theory in the 21 Century: applications for social justice studies.
In Denzin, N.K and Lincoln, Y.S., eds. (2005). The Sage Handbook of Qualitative Research, 3
Ed. California: Sage, pp.307-535.
Cheung, M.Y., Luo, C., Sia, C.L., and Chen, H. (2009). Credibility of Electronic Word-of-Mouth:
Informational and Normative Determinants of On-line Consumer Recommendations.
International Journal of Electronic Commerce. 13 (4), pp. 9–38.
Chorus, C., Molin E., Van Wee, B (2006a). Use and Effects of Advanced Traveller Information
Services (ATIS): a Review of the Literature. Transport Reviews, 26 (2), pp.127-149.
Chorus, C., Molin E., Van Wee, B (2006b). Travel information as an instrument to change cardrivers‟ travel choices: a literature review. European Journal of Transport and Infrastructure
Research. 6, (4), pp.335-364.
Christmas,S., Helman, S, Buttress,S., Newman, C. and Hutchins, R. (2010). Cycling, Safety and
Sharing the Road: Qualitative Research with Cyclists and Other Road Users. Simon Christmas
Ltd, Transport Research Laboratory and SHM Productions Ltd, Department for Transport,
Cialdini, R., Reno, R. and Kallgren, C. (1990). A Focus Theory of Normative Conduct: recycling
the concept of norms to littering in public places. Journal of Personality and Social Psychology.
58, pp.749-758.
Clifton, K.L. and Handy, S.L. (2003). Qualitative Methods in Travel Behaviour Research. In
Stopher, P. and Jones, P., eds. (2003). Transport Survey Quality and Innovation: Amsterdam:
Pergamon, pp. 283-302.
Cook, T. D., and Campbell, D.T. (1979) Quasi-Experimentation: Design and Analysis Issues for
Field Settings. Chicago: Rand McNally.
Couch, L.L. and Jones, W.H. (1997). Measuring Levels of Trust. Journal of Research in
Personality. 31, pp. 319-336.
Darley, J. M., & Latané, B. (1968). Bystander intervention in emergencies: Diffusion of
responsibility. Journal of Personality and Social Psychology, 8, pp.77-383.
Dellarocas, C. (2003). The Digitization of Word-Of-Mouth: Promise and Challenges of Online
Feedback Mechanisms. Management Science. 49 (10), pp.1407-1424.
Department for Transport, Local Government and the Regions (1998). The White Paper on
Transport. A New Deal Better for Everyone. Transport White Paper, TSO, London.
Department for Transport, Local Government and the Regions (2000). Transport 2010: the 10Year Plan. TSO, London.
Department for Transport, Local Government and the Regions (2002). Outline Transport Direct
Business Plan. London.
Department for Transport (2004). The Future of Transport. TSO, London.
Department for Transport (2005). Intelligent Transport Systems (ITS): the policy framework for
the roads sector, TSO, London.
Department for Transport (2011). Creating Growth, Cutting Carbon. Making Sustainable Local
Transport Happen. TSO, London.
Derrida, J. (1976). Of Grammatology. Baltimore, MD: Johns Hopkins University Press.
Deutsch, M. and Gerard, H.B. (1955). A study of normative and informational social influences
upon individual judgement. Journal of Abnormal and Social Psychology. 51, pp. 629-36.
Dholakia, U.M., Bagozzi, R.P. and Klein Pearo, L. (2004). A social influence model of consumer
participation in network- and small group-based virtual communities. International Journal of
Research in Marketing. 21, pp. 241 -263..
Eagly, A. and Kulesa, P. (1997). Attitudes, Attitude Structure and Resistance to Change:
Implications for Persuasion on Environmental Issues. In Bazerman, M.H., Messick, D.M.,
Tenbrunsel, A.E. and Wade-Benzoni, K.A., eds. (1997). Environment, Ethics and Behaviour: the
psychology of environmental valuation and degradation. San Francisco: New Lexington Press,
Eek, D., Loukopoulos, P., Fujii, S., and Gärling, T. (2002). Spill-Over Effects of Intermittent Costs
for Defection in Social Dilemmas. European Journal of Social Psychology. 32, pp. 801-813.
Esbjörnsson, M., Juhlin, J. and Östergren, M. (2004). Traffic encounters and Hocman:
Associating motorcycle ethnography with design. Personal Ubiquitous Computing. 8, pp. 92–99.
Farag. S. and Lyons, G. (2008a). What Affects Use of Pretrip Public Transport Information?
Empirical Results of a Qualitative Study. Transportation Research Record, 2069, pp.85–92.
Farag, S. and Lyons, G (2008b). Public transport information (non-)use empirically investigated
for various trips. Annual Meeting of the Transportation Research Board, January 2009,
Washington D.C.
Fehr, E. and Fischbacher, U. (2003). The nature of human altruism. Nature. 425, pp.785-791.
Festinger, L. (1950). Informal Social Communication. Psychological Review. 57, pp.271-82.
Fiedler, K. and Bless, B. (2001). Social Cognition. In Hewstone, M. and Stoebe, W., eds. (2001).
Introduction to Social Psychology: A European Perspective, 3rd Ed., Blackwell, pp.115-149.
Fincham, B. (2007). Bicycle Messengers: Image, Identity and Community. In Horton, D., Cox, P.
and Rosen P., eds.(2007), Cycling and Society, Chapter 9. Hampshire, UK: Ashgate Publishing
Ltd., pp.179-196.
Flyvbjerg, B. (2001). Making Social Science Matter: Why social inquiry fails and how it can
succeed again. Cambridge University Press.
Flyvbjerg, B. (2006). Five Misunderstandings about Case-Study Research. Qualitative Inquiry.
12 (2), pp. 219-245.
Fowler, J. H. and Christakis, N.A. (2009). Dynamic spread of happiness in a large social
network: longitudinal analysis over 20 years in the Framingham Heart Study. BMJ 2008;
Floridi, L. (2010), Information: A Very Short Introduction. Oxford University Press.
Freudendal-Pedersen, M., Hartmann-Petersen, K., and Drewes Nielsen, L. (2010). Mixing
Methods in the Search for Mobile Complexity. In Fincham, B., McGuinness, M. and Murray, L.,
eds. (2010), Mobile Methodologies. Palgrave Macmillan, pp.25-42.
Gabbott, M., & Hogg, G. (1994). Consumer behavior and services: A review. Journal of
Marketing Management. 10 (4), pp.311–324.
Gärling, T. (1998). Behavioural assumptions overlooked in travel-choice modelling. In J.d.D.
Ortuzar, D. A. Hensher, & S. Jara-Diaz, eds., (1998), Travel Behaviour Research: Updating the
State of Play. The Netherlands: Elsevier Science Ltd., pp.3-18.
Gärling, T. (1999). Value priorities, social value orientations and cooperation in social dilemmas.
British Journal of Social Psychology. 38, pp.397-408. .
Gatersleben, B., and Haddad, H. (2010). Who is the typical bicyclist? Transportation Research
Part F: Traffic Psychology and Behaviour. 13 (1), pp. 41-48.
Giddens, A. (1982). Profiles and Critiques in Social Theory. Berkeley: University of California
Giorgi, A. and Giorgi, B. (2008). Phenomenology. In Smith, J. A., (ed.). Qualitative Psychology:
A Practical Guide to Research Methods, Second Edition, Sage, London, pp.26-52.
Gilbert, D. C. (1991). Consumer behavior in tourism. In C. P. Cooper (Ed.) Progress in tourism,
recreation and hospitality management, Vol. 3 (pp. 78–105). Lymington, Hants, UK: Belhaven
Gitelson, R. L., and Crompton, J.L. (1983). The Planning Horizons And Sources of Information
used by Pleasure Travelers. Journal of Travel Research. 21, pp.2-7.
Gitelson, R., & Kerstetter, D. (1994). The influence of friends and relatives in travel decisionmaking. Journal of Travel and Tourism Marketing. 3 (3), pp.59–68.
Glaser, B.G. and Strauss, A.L. (1967). The Discovery of Grounded Theory: Strategies for
Qualitative Research. Aldine, Chicago
Godes, D., Mayzlin,D., Chen, Y., Das, S., Dellarocas, C., Pfeiffer, B., Libai, B., Sen, S., Shi, M.
and Verlegh, P. (2005). The Firm‟s Management of Social Interactions. Marketing Letters 16
(3/4), pp. 415–428.
Goffman, E. (1959). The Presentation of Self In Everyday Life. Gordon City, NY: Doubleday
and Co.
Government Social Research Unit (2007). The Magenta Book: Guidance Notes for Policy
Evaluation and Analysis. Background Paper 7: why do social experiments? London: HM
Granovetter, M. (1973). The Strength of Weak Ties. American Journal of Sociology. 78, (6)
Gretzel, U., Kyung, H.Y. and Purifoy, M.(2007). Online Travel Review Study, Laboratory for
Intelligent Systems in Tourism, Texas A&M University.
Hall, H. and Graham, D. (2004). Creation and recreation: motivating collaboration to generate
knowledge capital in online communities. International Journal of Information Management. 24
pp. 235-246.
Hall, H. and Widén-Wulff, G. (2008). Social Exchange, Social Capital and Information Sharing in
Online Environments: Lessons from Three Case Studies. In Huotari, M-L and Davenport, E. ,
From Information Provision to Knowledge Production. Proceedings of the international
conference for the celebration of the 20th anniversary of Information Studies, Faculty of
Humanities, June 23–25 2008, University of Oulu, Finland.
Halpern, D., Bates, C., Mulgan, G. and Aldridge, S., with Beales G. and Heathfield, A. (2004):
Personal Responsibility and Changing Behaviour: the state of knowledge and its implications for
public policy, Prime Minister‟s Strategy Unit, London.
Hammersley, M. (1992). What‟s Wrong With Ethnography? London: Routledge.
Hewstone, M. and Stoebe, W., eds. (2001). Introduction to Social Psychology: A European
Perspective, 3rd Ed. Malden, MA: Blackwell.
Homans, G.C. (1958). Social Behaviour as Exchange. The American Journal of Sociology. 63
(6), pp.597-606.
Honeycutt, C. and Herring, S.C. (2009). Beyond Microblogging: conversation and collaboration
via Twitter. Proceedings of the Forty-Second Hawaii International conference on System
Sciences (HICSS-42). Los Alamitos, CA: IEEE Press.
Husserl, E. (1936, trans.1970). The Crisis of European Sciences and Transcendental
Phenomenology. Euanston: Northwestern University Press.
Jackson, P.G. and Polak, J.W. (1997). Behavioural responses to information: a review. Centre
for Transport Studies, Imperial College, London.
Jackson, T. (2005) Motivating Sustainable Consumption: a review of evidence on consumer
behaviour and behavioural change. Report to the Sustainable Development Research Network.
Jansen, B.J., Zhang, M., Sobel, K. and Chowdury, A. (2009). Twitter Power: Tweets as
electronic word of mouth. Journal of the American Society for Information Science and
Technology. 60 (11), pp.2169-2188.
Johnson-George, C. and Swap, W.C. (1982). Measurement of Specific Interpersonal Trust:
Construction and Validation of a Scale to Assess Trust in a Specific Other. Journal of Personality
and Social Psychology. 43 (6), pp.1306-1317.
Jones, P.M., Dix, M.C., Clarke, M.I. and Heggie, I.G. (1983). Understanding Travel Behaviour.
Aldershot, England: Gower Publishing Company Ltd.
Karl, C.A. and Bechervaise, N.E. (2003). The learning driver: issues for provision of traveller
information services. 10 World Congress and Exhibition on Intelligent Transport Systems and
Services, Madrid.
Kelley, H.H. and Thibaut, J.W. (1978) Interpersonal Relations: a theory of interdependence,.
New York: Wiley.
Kenyon, S., Lyons, G. (2003). The value of integrated multimodal traveller information and its
potential contribution to modal change. Transportation Research Part F. 6 (1), pp 1-21.
Kozuki, A. and Mahmassani, H. S.(2009). Information acquisition and social interaction
mechanisms in opinion formation and market adoption of transportation services. Annual
Meeting of the Transportation Research Board, January 2009. Washington D.C.
Kozinets, R.V. (1999). E-Tribalised Marketing? The Strategic Implications of Virtual
Communities of Consumption. European Management Journal. 17 (3), pp.252-64.
Kozinets, R.V. (2010). Netnography: Doing Ethnographic Research Online. London: Sage.
Kramer, R. M. & Brewer, M. B. (1984). Effects of group identity on resource use in a simulated
commons dilemma. Journal of Personality and Social Psychology. 46 (5), pp1044-1057.
Kruegar, R.A. (1994). Focus Groups : a practical guide for applied research. California: Sage.
Lamont, D. and Lyons, G. (2007). Understanding and addressing dyslexia in travel information
provision. Proceedings of the 39 annual UTSG conference, January, 2007, Harrogate, UK.
Latané, B, & Nida, S. (1981). Ten years of research on group size and helping. Psychological
Bulletin, 89, 2, pp. 308-324.
Leary, M.R. (1995). Self-Presentation. Impression Management and Interpersonal Behaviour.
Oxford: Westview Press.
Lewis, J. and Ritchie, J. (2003). Generalising from Qualitative Research. In Ritchie, J. and
Lewis, J. (eds) Qualitative Research Practice: a guide for social science students and
researchers. London: Sage.
Liebrand, W. (1984). The effect of social motives, communication and group size on behaviour in
an N-person multi-stage mixed motive game. European Journal of Social Psychology. 14,
Line, T.,Jain, J. and Lyons, G.(2010). The role of ICTs in everyday mobile lives. Journal of
Transport Geography [online]. DOI:10.1016/j.jtrangeo.2010.07.002.
Ling, K., Beenen, G., Ludford, P., Wang, X., Chang, K., Li, X., Cosley, D., Frankowski, D.,
Terveen, L., Rashid, M., Resnick, P., Kraut, R. (2005). Using Social Psychology to Motivate
Contributions to Online Communities. Journal of Computer-Mediated Communication [online], 10
(4). DOI: 10.1111/j.1083-6101.2005.tb00273.x
Lyons, G. (2006). The role of information in decision-making with regard to travel. Science
Review for Foresight Intelligent Infrastructure Systems Project, Department for Trade and
Industry, London.
Lyons, G., Avineri, E., Farag, S. and Harman, R. (2007). Strategic Review of Travel Information
Research. Department for Transport, London.
Lyons, G., Goodwin, P., Hanly, M., Dudley, G., Chatterjee, K., Anable, K, .Wiltshire, P., and
Susilo, Y. (2008). Public Attitudes to Transport: Knowledge Review of Existing Evidence. Final
Report to the Department for Transport, London.
Mann, C. and Stewart, F. (2000). Internet Communication and Qualitative Research: A
Handbook for Researching Online. London: Sage.
Mason, J. (2002). Qualitative Researching (2 ed.). London: Sage.
McAllister, D.J., (1995). Affect- and cognition-based trust as foundations for interpersonal
corporation in organisations. Academy of Management Journal. 38, pp. 24-59.
McClintock, C.G. (1972). Social motivation – a set of propositions, Behavioural Science. 17 (5),
pp. 438-454.
Mead, G.H. (1934). Mind, Self and Society. Chicago: University of Chicago Press.
Mendes-Fihlo, L. and Tan, F.B. (2009). User-Generated Content and Consumer Empowerment
in the
Travel Industry: A Uses & Gratifications and Dual-Process Conceptualization (2009).
Pacific-Asia Conference on Information Systems 2009 Proceedings. Paper 28. 10-12 July, 2009,
Hyderabad, India.
Miles, M.B. and Huberman, A.M. (1994). Qualitative Data Analysis: An Expanded Sourcebook,
2 ed. Thousand Oaks, CA: Sage.
MORI (2001). The CfIT Report 2001: Public Attitudes to Transport in England. Commission for
Integrated Transport, London.
Murphy, L., Mascardo, G. and Brenckendorff, P. (2007). Exploring word-of-mouth influences on
travel decisions: friends and relatives vs. other travellers. International Journal of Consumer
Studies 31, pp. 517-527.
Musselwhite, C.B.A., Avineri, E., Susilo, Y.O., Fulcher, E., Bhattachary, D., Hunter, A. and
Stockley, R. (2010). Understanding the public attitudes to road user safety: Final report. Road
safety research report No.111. Department for Transport, London.
North Jersey Transportation Planning Authority (NJTPA) and New Jersey Institute of Technology
(NJIT) (2010). Transportation Agencies Find New On-Ramps to the Information Superhighway.
Intransition: Transportation Planning, Practice and Progress. 17 (1).
Oakley, A. (1981). Interviewing women: a contradiction in terms. In Roberts, H., ed. (1981),
Doing Feminist Research. London: Routledge and Kegan Paul plc., pp.30-61.
Ormerod, P. (2010). N Squared. Public policy and the power of networks. Royal Society of Arts,
Essay 3, August 2010.
Owens, S. (2000) Commentary: Engaging the public: information and deliberation in
environmental policy. Environment and Planning A. 32 (7), pp.1141-1148
Patton, M.Q. (2002). Qualitative Research and Evaluation Methods, 3 ed., Thousand )aks, CA:
Payne, J.W., Bettman, J.R. and Johnson, E.J. (1993). The Adaptive Decision Maker. Cambridge
University Press.
Perugini, M. and Conner, M. (2000). Predicting and understanding behavioural volitions: the
interplay between goals and behaviours. European Journal of Social Psychology. 30 (5), pp.705731.
Peters, O. and ben Allouch, S. (2005). Always connected: A longitudinal field study of mobile
communications, Telematics & Informatics. 22 (3), pp.239-256.
Priedhorsky, R., Jordan, B. and Terveen, L. (2007). How a personalized geowiki can help
bicyclists share information more effectively. Geowiki ‟07: Proceedings of the 2007 international
symposium on Wikis, ACM, New York.
Pruitt, D.G. and Kimmel , M.J. (1977). Twenty years of experimental gaming : critique, synthesis,
an suggestions for the future. Annual Review of Psychology. 28, pp.363-92.
Putnam, D.R (2000) Bowling Alone: the Collapse and Revival of American Community. New
York: Simon & Schuster.
Rheingold, H. (1993). The Virtual Community: Homesteading on the Electronic Frontier [online].
Reading, MA: Addison-Wesley. http://www.rheingold.com/vc/book/3.html. [Accessed 25/06/10].
Ritchie, J. and Lewis, J. (eds) (2003). Qualitative Research Practice. London: Sage.
Ritchie, J., Spencer, L. and O‟Connor, W. (2003). Carrying out Qualitative Analysis. In Ritchie, J.
and Lewis, J., eds. (2003), Qualitative Research Practice: a guide for social science students
and researchers. London: Sage Publications, pp.219-262.
Robson, C. (1993). Real World Research: a resource for social scientists and practitioner
researchers. Oxford: Blackwell.
Robson, C. (2002). 2
ed. Real World Research: a resource for social scientists and practitioner
researchers. Oxford: Blackwell.
Rogers, E. (2003) Diffusion of Innovations, Free press, New York
Rousseau, D.M, Sitkin, S.B., Burt, R.S. and Camerer, C. (1998). Not so Different After All: a
Cross-discipline View of Trust. Academy of Management Review. 23 (3), pp.393-404.
Schwartz, S.H. (1977). Normative influences on altruism. In Berkowitz, L. (Ed.) Advances in
Experimental Social Psychology 10, New York, Academic Press
Schwartz, S.H. (1994). Are there universals in the structure and content of human values?
Journal of Social Issues. 50 (4), pp.19-45.
Simon, H. (1959). Theories of Decision-Making in Economics and Behavioral Science. American
Economic Review. 49, pp.253-283.
Sirakaya, E. and Woodside, A.G. (2005). Building and testing theories of decision making by
travellers. Tourism Management. 26, pp.815–832
Skinner, D. and Rosen, P. (2007). Hell is other Cyclists: Rethinking Transport and Identity. In
Horton, D., Rosen, P. and Cox, P. , eds. (2007). Cycling and Society. Hants, UK: Ashgate,
Smith, J.A., and Osborn, M. (2008). Interpretive Phenomenological Analysis. In Smith, J. A., ed.,
(2008). Qualitative Psychology: A Practical Guide to Research Methods, 2nd Ed. London: Sage,
Snape, D. and Spencer, L. (2003). The Foundations of Qualitative Research. In Ritchie, J. and
Lewis, J., eds. (2003). Qualitative Research Practice, a guide for social science students and
researchers. London: Sage, pp.1-23.
Social Research Associates (2004). Transport Direct Market Research - MR07- Incorporating
Cost, Reliability and other Travel Factors into Journey Planning. Department for Transport,
Social Research Associates (2005). Transport Direct Market Research – Stage 2b: How People
Choose. Department for Transport, London.
Spencer, L., Ritchie, J., and O‟Connor, W. (2003). Analysis: Practices, Principles and
Processes. In Ritchie, J. and Lewis, J., eds. (2003). Qualitative Research Practice: a guide for
social science students and researchers. London: Sage, pp.199-218.
Spinney, J. (2009a). Mobilising sustainability: partnership between a pro-cycling NGO and local
government in London (UK). The 6 Cycling and Society Research Group Symposium,
September 2009, Bolton, UK.
Spinney, J. (2009b). Cycling the City: Movement, Meaning and Method. Geography Compass
3/2 , pp.817-835.
Staats, H., Harland, P. and Wilke, H.A.M. (2004). Effecting Durable Change: A Team Approach
to Improve Environmental Behavior in the Household. Environment and Behavior. 36, pp.341367.
Steg, L., Vlek, C., & Slotegraaf, G. (2001). Instrumental-reasoned and symbolic-affective
motives for using a motor car. Transportation Research F. 4, pp.151–169.
Stigler, G.J. (1961). The economics of information. Journal of Political Economy. 69, pp. 213225.
Strauss, A. and Corbin, J. (1990). Basics of Qualitative Research: grounded theory procedures
and techniques. London: Sage.
Stroebe, W. and Jonas, K. (2001). Health Psychology: a social-psychological perspective. In
Hewstone, M. and Stroebe, W., eds (2001), Introduction to Social Psychology: a European
Perspective. Blackwell, pp.519-553.
Sunitiyoso, Y. (2008). Role of Social Interaction, Social Learning and Social Influence in the
Dynamics of Travellers‟ Mode Choice Behaviour. PhD Thesis, University of the West of England,
Tajfel, H. (1982). Social Psychology of Intergroup Relations. Annual Review of Psychology. 33,
Tajfel, H. and Turner, J. (1986). The social identity theory of intergroup behaviour. In S. Worchel
and W.G. Austin, eds. (1986), Psychology of intergroup relations, Chicago: Nelson-Hall, pp.7-24.
Taniguchi, A., Hara, H., Takano, S., Kaygaya, S. and Fujii, S. (2003). Psychological and
Behavioral Effects of Travel Feedback Program for Travel Behavioral Modification. Annual
Meeting of the Transportation Research Board, January 2003, Washington D.C.
Taniguchi, A. and Fujii, S. (2007). Promoting public transport using marketing techniques in
mobility management and verifying their quantitative effects. Transportation (2007) 34 (1), pp.3749.
Tapp, A., Spotswood, F. and Leonard, S. (2010). The image of cycling in Britain. Presentations
at Understanding and Motivating Cycling seminar, December 2010, University of the West of
England, Bristol.
Taylor, M.A.P, and Ampt, E.S. (2003).Travelling smarter down under: policies for voluntary
behavioural change in Australia. Transport Policy. 10 (3), pp.165-177.
Terry, D., Hogg, M. and White, K (1999). The theory of planned behaviour: Self-identity, social
identity and group norms. British Journal of Social Psychology (1999), 38, 225-244.
Thaler, R.H. and Sunstein, C.R. (2009). Nudge. Yale University Press.
Transportation Research Group (2000). Establishing User Requirements from Traveller
Information Systems. Final report to the Engineering and Physical Sciences Research Council,
University of Southampton.
Triandis, H. (1977) Interpersonal Behaviour. Monterey: Brooks and Cole.
Triandis, H. (2001). Individual-Collectivism and Personality. Journal of Personality. 69 (6),
TTR (2004). Traveline and Transport Direct disabilities customer research. Department for
Transport, London.
Trivers, R. L. (1971). Evolution of Reciprocal Altruism. Quarterly Review of Biology. 46 (1), pp.
Tu, C.H. and McIsaac (2002). The Relationship of Social Presence and Interaction in Online
Classes. The American Journal of Distance Education. 16 (3), pp.131-150.
Turner, J.C., Hogg, M.A., Oakes, P.J., Reicher, S.D. and Wetherell, M.S.(1987). Rediscovering
the Social Group. A Self-Categorisation Theory. Oxford: Blackwell.
Turner, J.C. (1991). Social Influence. Buckingham: Open University Press.
Turner, J.C., Oakes, P.J., Alexander Haslam, S. and McGarty, C. (1994). Self and Collective:
Cognition and Social Context. Personality and Social Psychology Bulletin. 20 (5), pp. 454-463.
Tversky, A and D Kahneman 1974. Judgement under uncertainty: heuristics and biases.
Science 185 (4157), pp,1124-1131.
Um, S. and Crompton, J.L (1990). Attitude Determinants in Tourism Destination Choice. Annals
of Tourism Research. 17, pp. 432-448.
Urry, J. (2000) Sociology Beyond Societies. London:Routledge.
Urry, J. (2002) Mobility and Proximity. Sociology, 36 (2), pp. 225-274
Urry, J. (2004). Small Worlds and the New „Social Physics‟. Global Networks. 4 (2), pp.109-30.
Van der Horst, R. (2006). Getting there and away: the role of travel information in recreational
day trips, with a specific focus on the mode and the destination choice. PhD thesis. Utrecht
Van Lange, P., van Vugt, M., Meertens, R., and Ruiter, R. (1998). A Social Dilemma Analysis of
Commuter Preferences: The Roles of Social Value Orientation and Trust. Journal of Applied
Social Psychology. 28 (9), pp.796-820.
Van Lange, P., and De Dreu, C. (2001). Social Interaction: Cooperation and Competition. In
Hewstone, M., Stroebe, W., eds. (2001), Introduction to Social Psychology, Oxford: Blackwell
Publishing, pp.341-369.
Van Vugt, M., Meertens, R., van Lange P. (1995). The role of social value orientations in a reallife social dilemma. Journal of Applied Social Psychology. 25 (3), pp.258-278.
Van Vugt, M. (1997). Concerns about the Privatisation of Public Goods: a Social Dilemma
Analysis. Social Psychology Quarterly. 60 (4), pp.355-367.
Verplanken, B., Aarts, H., van Knippenberg, A. (1997). Habit, information acquisition and the
process of making modal choices. European Journal of Social Psychology. 27, pp.539-560.
Walter, C. K., and Tong, H.M. (1977). A Local Study of Consumer Vacation Decisions. Journal of
Travel Research. 15, pp.30-34.
Walther, J.B. (1992). Interpersonal effects in computer-in mediated interaction: A relational
perspective. Communication Research. 19, pp.52-90.
Walther, J.B. (1995). Relational Aspects of Computer-mediated Communication: Experimental
Observations over Time. Organization Science. 6 (2), pp.186-203.
Wetherell, M.S. (Ed.) (1996). Identities, Groups and Social Issues. The Open University.
London: Sage.
Wilson, J.M., Straus, S.G. and McEvily, B. (2006). All in due time: The development of trust in
computer-mediated and face-to-face teams. Organizational Behavior and Human Decision
Processes.99, pp.16-33.
Yin, R.K. (2009). Case-study research: design and methods (fourth ed.). London: Sage.
Ziebland, S., Chapple, A., Dumelow, C., Evans, J., Prinja, S.,and Rozmovits, L. (2004). How the
internet affects patients‟ experience of cancer: a qualitative study. BMJ. 2004 328:564.
Zhang, W. and Watts, S. (2008). Online communities as communities of practice: a case-study.
Journal of Knowledge Management. 12 (4), pp.55-71.
Appendix A – Phase 1 Interview Topic Guide
1. Introduction
Introduce self, and tell of my position.
Outline PhD research project:
My research is about the use of different types of travel information, including informal
information obtained through word-of-mouth, and its influence on peoples‟ decisions about how,
when and where to travel. One of the ways I‟d like to explore this is by asking people how they
found out about ways of getting to UWE when they first started here, as well as their use of
different types of travel information in other circumstances.
Purpose of interview research
Overview of what the interview will cover
The interview will take a maximum of 1 hour.
Material used from the interview will be used in the PhD thesis – quotes from
interviewees may be used, and these will be non-traceable and completely
anonymous  it may say things like “Male”.
Right to leave at any time during the interview
No right or wrong answers – just want thoughts and opinions. If there are any questions
you prefer not to answer, you are free to do so.
Any questions?
Are you still happy for the interview to be recorded? (this will have been raised
previously when interview was arranged)
Let‟s start.
2. Warm-up information: experiences, circumstances, behaviours
These questions are intended to be easy to answer, in order to break the ice, provide some
contextual information about the interviewee, and introduce the concept of formal and
informal travel information. Replies will not be probed. Spend a maximum of 10 minutes on
this section.
2.1 About yourself
Please tell me a little about yourself in relation to UWE:
Are you student/staff, full-time/part-time?
How long have you been at UWE?
Which campuses do you go to, and how often?
Which area do you live in (how far do you need to travel)?
2.2 Your travel behaviour
How do you usually travel to and from UWE?
If by car, do you car-share, and if so, how often?
What other modes of transport, if any, do you sometimes use to get to or from UWE?
In general (aside from your journey to/from UWE) what forms of transport do you tend to
2.3 Your use of travel information
(Travel information can include information on, for example, travel times, costs, routes, or more
personal advice relating to comfort, convenience, safety etc. It might be obtained from formal
sources such as paper timetables, internet journey planners, telephone enquiries, or informal
sources such as friends and family).
For what sort of trips do you tend to use travel information?
Prompts: different trip purposes or modes of travel
Business, leisure, shopping, holiday, going to an event, taking children
Different modes of travel
Generally speaking, what sources of travel information do you use if you need to plan an
unfamiliar trip?
Prompts: formal and informal sources.
Generally speaking, what types of travel information do you use if you need to plan an
unfamiliar trip?
Prompts: formal and informal types.
3. Core questions: attitudes and explanation
The first two sets of questions in this section are intended to identify a trip example which will
yield some in-depth reflection on use and influence of informal travel information..
3.1 Remembering when you first travelled to UWE:
How did you find out about ways of getting to and from UWE? (prompts: “ways” can
include transport mode, routes, times etc).
If interviewee mentions only formal information, prompt with:
When you first started here, did you ask for, or were you offered, any informal
information or advice from other people about travelling to UWE?
If so, please tell me a little about this information. How far, if at all, do you think it
influenced your decisions about how/when to travel?
If not, why not?
If the response to this question offers enough scope to probe about informal information,
proceed straight to 3.3. If not, ask:
Have you received any informal information about travel to/from UWE since you first
started here?
If so, please tell me a little about this information. How far, if at all, do you think it
influenced your decisions about how/when to travel?
If not, why not?
If the response offers enough scope to probe, proceed to 3.3. if not, ask:
3.2 Thinking about other sorts of trip:
Can you think of an example of another trip you needed to plan? (prompts: e.g.
business, leisure, shopping, holiday, going to an event, taking children somewhere).
If so, please give details:
where you were going and why
whom you were travelling with
by what mode of transport
What sort of travel information did you use to help plan this trip?
Did you receive any informal information or advice from other people?
If so, please tell me a little about this information. How far, if at all, do you think it
influenced your decisions about how/when to travel?
If not, why not?
If not, ask whether they can think of an example when they did.
3.3 The use and influence of informal information (compared with formal information) for
this particular trip
I‟d like to hear in more detail about the informal information you have used for your travel to
UWE/other trip example, compared with formal information.
Who was the other person/people who gave you this information/advice?
Who else might you have asked and why?
Attitudes to/relationship with information–giver
How much confidence did you have in the information/advice you received?
Why was this the case? Did you trust it more or less than formal sources of
information? (Trust)
How much did you feel you identified with the person/people giving you the
information/advice? (Social Identity). What kind of things do you have in
common with them (e.g. gender, age, occupation etc?)
Is the person/people who gave you the information particularly important to you
(e.g.close friend/family member/partner)? If so, how important was it to you that
they approved of your transport arrangements for this particular trip?
(Subjective norm)
What advice did they give you?
Did you act on it?
Did you ask them for this information, or did they offer it voluntarily?
If you asked them, why did you want/need this informal information?
Prompts: saving time/effort; information not available from formal sources;
opportunity for social interaction.
What do you think you learned which you think you could not have learned from formal
types and sources of information? Prompts: routes, services, times, convenience,
reliability, parking etc.
Attitudes to information content (e.g. transport mode) :
To what extent did you act on this advice because it related to “normal” travel
behaviour (something everyone else does), or a behaviour you consider to be
the “right” thing to do (something everyone ought to do)? (Social norm)
Why do you think they provided you with it?
How did these discussions take place? Prompts: face-to-face, by phone/email etc…
How carefully did you feel you needed to plan this journey? How important was other
people‟s advice to you in reducing uncertainties? (e.g. risk of being late, getting lost etc?
How far could you have planned this trip on your own without asking others for advice?
How do you feel, generally, about asking people for advice (self-presentation)
- If you don‟t ask people, why not?
How typical is the example you have just described of your use of travel information
more generally?
3.4 Turning these questions around to information you have given to others:
What type of information/advice have you given to other people about travelling to/from
UWE? (if interviewee can‟t think of examples, ask about another specific trip)
To whom did you give this information/advice?
(prompt: friends, other students on your course, colleagues, strangers etc)
Did they ask you, or did you offer your advice without being asked?
Why did you respond to their request, or offer this advice?
Expect interview to say they wanted to be helpful, but try to probe reasons, e.g.
pro-social values
conforming with social norms (expected behaviour to assist someone when
they ask)
wanting to look helpful (self-presentation)
wanting to help this person because he/she is like me (social identity)
Wanting to influence the other person towards a particular travel behaviour (e,g.
because you feel strongly about it).
In what ways do you think this information might have influenced their travel choices?
4. Wind down: suggestions/looking to the future
Can you think of any improvements which could be made to the travel information
available to people coming to UWE?
In what ways might informal information be made widely available to people coming to
UWE (or for transport more generally)?
5. Close
That is the end of the interview
Do you have any questions or anything you would like to add to what you have said
Thank you for taking part in the interview
Appendix B – Phase 1 Focus Group Topic Guide
Before beginning the discussion: offer refreshments
Place name cards on table.
1. Scene-setting and ground rules:
Introduce self and research
My research is about the use of different types of travel information, including informal
information obtained through word-of-mouth, and its influence on peoples‟ decisions about how,
when and where to travel. One of the ways I‟d like to explore this is by asking people how they
found out about ways of getting to UWE when they first started here, as well as their use of
different types of travel information in other circumstances.
Housekeeping (toilets, exit?)
Expected roles of participants. No right or wrong answers – just want thoughts and
opinions. If there are any questions you prefer not to answer, you are free to do so.
Any questions?
Recording of discussion (please only one person speaking at once), note-taking
Confidentiality: Material used from the interview will be used in the PhD thesis – quotes
from interviewees may be used, and these will be non-traceable and completely
Consent forms
Let‟s start. Switch on tape recorder
2. Individual introductions
Ask each member of the group to introduce themselves:
Area of study
Jot down table plan with names.
3. Opening topics (10 minutes):
Your travel behaviour
By what means do you usually travel to and from UWE (car etc……)?
Does anyone ever use any other modes for this trip?
Which areas do you live in (how far do you need to travel)?
In general (aside from your journey to/from UWE) what forms of transport do
you tend to use?
Your use of different types of travel information
I‟d like to move on now to the topic of travel information, which is the subject of this
study. Here are a few pictures, illustrating some sources of travel information.
What do you think of these examples? Can you suggest any others?
Which of these do you tend to use?
For what sort of trips would you use them? E.g. more for some modes than
Try to cover both pre-trip and on-trip.
What type of information do you tend to receive or pass on through word of
mouth? (e.g. convenience, comfort, reliability)
Does anyone use any informal, electronic means of information-sharing
concerning travel (e.g. plotting routes on maps, social networking, email
discussion forums, message boards)
4. Discussion
4.1 Remembering the first time you came to UWE (20 minutes)
NB need to work hard at steering the discussion away from general comments about transport
services and road congestion.
How did you find out about ways of getting to and from UWE during your first two
weeks? (prompts: “ways” can include transport mode, routes, times etc).
When you first started here, did you ask for, or were you offered, any informal
information or advice from other people about travelling to UWE?
If so, please tell me a little about this information. How far, if at all, do you think it
influenced your decisions about how/when to travel?
If not, why not?
Did it just come up in conversation?
Do you think you have influenced anyone else by passing on informal information?
If so, how?
If not, why not?
4.2 Example of another unfamiliar trip (20 minutes)
Ask everyone to think of an unfamiliar trip they have made in the past year. This might be, for
example, a trip for work, leisure, shopping, a holiday, going to a special event, or travelling with
someone they wouldn‟t normally go to this place with.
Please give details of the trip:
What sort of travel information did you use to help plan this trip?
Did you seek any informal information or advice from other people?
If yes, what, and from whom?
If not, why not?
did it just come up in conversation?
How far do you think your decision about how, when or where to travel was influenced
by informal information?
How did it influence you (e.g. mode choice, route)?
4.3 Reasons why you might have been influenced by informal information (30 minutes)
Use sort cards demonstrating some social-psychological factors: get group to organise them in
order of importance and discuss choices. Have blank cards so that other factors can be
identified during the discussion.
Attitudes to/relationship between information–giver and recipient
- The person was a close friend/family member.
- I trusted the other person.
- I felt I had things in common with the other person.
- I wanted the other person to approve of my travel arrangements.
Attitudes to information content
- I had confidence in this information.
- the information I was given related to "normal" travel choices (most people would travel in this
- the information I was given related to travel choices, which I think people should make
Attitudes to seeking and acting on information (differentiate between pre-trip and on-trip)
- I am generally happy to approach other people for advice.
- I needed to plan this journey carefully
- other people's advice was important to me to reduce the uncertainty of this journey.
- I could not have planned this trip without asking someone's advice.
4.4 Reasons why you have given information to others (differentiate between pre-trip and
- I wanted to make their life easier.
- I was confident about the accuracy of the advice I was giving
- I am a helpful person.
- I wanted to influence the other person to make travel choices, which I believe are right.
- it is "normal" to help people.
- I wanted to look helpful.
- I wanted to help the other person because they are like me
- It just came up in the conversation
5. Winding down (10 minutes)
Ending the discussion: signal that end of discussion is approaching.
Do you think it might be helpful if informal information were more widely available to
people coming to UWE (or for transport more generally)?
If so, how could this be done?
Would anyone like to say anything else?
Thank you for taking part.
Give out the money/vouchers (recipients must sign to say they have received it).
Appendix C – Exploring social-psychological factors:
examples from research
1. Role beliefs (TIB)
Bamberg, B. and Schmidt, P. (2003). Incentives, Morality of Habit? Predicting Students‟ Car
Use for University Routes With the Models of Ajzen, Schwartz and Triandis. In Environment
and Behaviour 2003; 35; 264. Sage Publications.
For me as a student it is (appropriate/not appropriate) to use the car for university
Using a car for university routes is (fitting/not fitting) my position as a student.
NB Role beliefs are linked to self-identity (identity as a student, academic, mother etc).
2. Subjective norm (TPB)
Bamberg and Schmidt (ibid).
How strong would 1) your friends; 2) your partner support you if you use the car for
university routes next time?
Perugini, M. and Connor, M. (2000). Predicting and understanding behavioural volitions: the
interplay between goals and behaviours. In European Journal of Social Psychology, 30, 705731. John Wiley and Sons ltd.
People who are important to me think I should/should not perform activity Y in the next 4
weeks to try to achieve goal X.
People who are important to me would approve/disapprove of my performing activity Y
in the next four weeks to try to achieve goal X.
People who are important to me want me to perform activity Y in the next four weeks to
try to achieve goal X.
3. Personal norm (Schwartz‟s norm activation model)
Bamberg and Schmidt (ibid).
If I use the car for my university routes next time I would have a moral stomachache (likely/unlikely)
Not using environmentally friendly travel modes like bike or public transport for
university routes next time would violate my principles (agree/disagree)
How strongly do you fell a personal obligation to use environmentally friendly
travel modes like a bike or public transport for university routes next time?
4. Social Norms
Sherif, M. (1936) The Psychology of Social Norms. New York. Harper and Brothers (Harper
Torchbook edition, 1966).
A very broad area! Social norms are described by Sherif (1936) as “customs, traditions,
standards, rules, values, fashions and all other criteria of conduct which are standardised as
a consequence of the contact of individuals.” (1936, p3). He conducted experiments to test
the process of consensus formation within groups, showing how shared social norms
Injunctive and descriptive social norms (and personal norms) were tested in an experimental
field setting (observing whether people dropped litter under controlled conditions by:
Cialdini, Robert, R Reno and C Kallgren 1990. A Focus Theory of Normative Conduct:
recycling the concept of norms to littering in public places. Journal of Personality and Social
Psychology 58, 749-758.
However, “There is no definitive scientific theory of social norms” (Turner, J.C., 1991, p2).
Turner, J.C. (1991) Social Influence. Open University Press, Buckingham
5. Prestige
Ellaway, A., Macintyre, S., Hiscock, R. and Kearns, A. (2003). In the driving seat:
psychosocial benefits from private motor vehicle transport compared to public transport.
Transportation Research Part F. 6, 217-231
Most people would like to travel by the type of car/public transport that I use
When I travel by car/public transport it makes me feel I‟m doing well in life
Cramer, K., and Gruman, J. (2002). The Lennox and Wolfe revised self-monitoring scale:
Latent structure and gender invariance. Personality and Individual Differences, 32(4), 627637
Lennox and Wolfe (1984) Revised Self-Monitoring Scale
In social situations, I have the ability to alter my behavior if I feel that something else is
called for.
I am often able to read people‟s true emotions correctly through their eyes.
I have the ability to control the way I come across to people, depending on the
impression I wish to give them
In conversations, I am sensitive to even the slightest change in the facial expression of
the person I‟m conversing with.
My powers of intuition are quite good when it comes to understanding others‟ emotions
and motives.
I can usually tell when others consider a joke to be in bad taste, even though they may
laugh convincingly.
When I feel that the image I am portraying isn‟t working, I can readily change it to
something that does.
I can usually tell when I‟ve said something inappropriate by reading it in the listener‟s
I have trouble changing my behavior to meet the requirements of any situation I find
myself in.
I have found that I can adjust my behavior to meet the requirements of any situation I
find myself in.
If someone is lying to me, I usually know it at once from that person‟s manner of
Even when it might be to my advantage, I have difficulty putting up a good front
Once I know what the situation calls for, it‟s easy for me to regulate my actions
7. Self-identity (Stryker)
Sparks, P. & Shepherd, R. (1992). Self-identity and the theory of planned behavior:
Assessing the role of identification with "green consumerism". Social Psychology Quarterly,
55(4), 388-399.
I think of myself as a 'green consumer';
I think of myself as someone who is very concerned with 'green issues'.
Terry, D.J., Hogg, M.A. and White, K.M. (1999). The theory of planned behaviour: Selfidentity, social identity and group norms. British Journal of Social Psychology, 38, 225-244
`to engage in household recycling is an important part of who I am‟ (1 5 no,
definitely not to 7 5 yes, definitely) ;
I am not the type of person oriented to engage in household recycling ‟ (1 5
completely false to 7 5 completely true) ;
` I would feel at a loss if were forced to give up household recycling ‟ (1 5
strongly disagree to 7 5 strongly agree)
Self identity is linked to role beliefs.
8. Social Identity (in-group, out-group – Tajfel)
NB. Both self-identity and social identity contribute to self concept. In social identity theory,
identity emanates from the group memberships.
Terry, D.J., Hogg, M.A. and White, K.M. (1999). The theory of planned behaviour: Selfidentity, social identity and group norms. British Journal of Social Psychology, 38, 225-244
Group norms:
How many of your friends and peers would engage in household recycling ? ‟ (1 5 none
to 7 5 all)
`Most of my friends and peers think that me engaging in household recycling during the
next fortnight would be..‟ (1 5 undesirable to 7 5 desirable).
Identification with the reference group:
`How much do you identify with your group of friends and peers? ‟ ; 1 5 not very much to
7 5 very much)
` In general, how well do you feel you fit into your group of friends and peers ? ‟ ; 1 5 not
very much to 7 5 very much.
9. Social Values
Liebrand, W: The ring measure of social values.
How would you share a resource between yourself and an unknown other?
10. Social influence
Yos Sunityoso: draft thesis
Would the number of people currently participating in car sharing affect your decision on
whether or not to join the programme?
For example: You may consider joining car sharing if a significant number of people
have already been joining.
If your answer is Yes, what is the percentage of students at Frenchay campus
participating in car sharing programme that may encourage you to join?
11. Homophily/heterophily
These concepts, developed by sociologists in the 1950s, were used by Everett Rogers in The
Diffusion of Innovations (1983); also known as the “Like me principle” - Laumann, E. O. (1966),
Prestige and Association in an Urban Community, Indianapolis, IN: Bobbs-Merrill.
Hypotheses about the role of homophily and heterophily in WOM referral behaviour were tested
in the following study:
Johnson Brown, J. and Reinigen, P.H. (1987). Social Ties and Word-of-Mouth Referral
Behaviour. In The Journal of Consumer Research, Vol. 14, No.3 (Dec., 1987), pp350-362.
In this strong and weak ties analysis, standard socio-demographic questions were asked about
respondents‟ occupation, age, gender, education etc. Ties were classed as homophilous
(heterophilous) when they had identical (different) category memberships for at least three of
these attributes.
Appendix D – The influence of theory-driven analysis
Phenomenological analysis
Phenomenological psychological research involves investigating and analysing lived examples
of the phenomenon under study within the context of the participants‟ lives, in order to
understand the psychological meanings that constitute the phenomenon. Phenomenological
techniques can be used to reveal meanings of which the research participants themselves may
not be aware. Giorgi and Giorgi (2008) state that: "while persons‟ awarenesses are concomitant
with these lived experiences, they are hardly ever totally coincident to what is being experienced
by them. Usually, the capacity to live through events or respond to different situations greatly
exceeds the capacity to know exactly what we do or why we do what we do." (pp28). Arguably,
such attempts to read below the surface of participants‟ accounts are common to any
interpretive reading of qualitative data, but are particularly useful when, as is the case in this
PhD, underlying psychological meanings are sought.
The "database" in phenomenology is often based on retrospective descriptions, since the key
data to be analysed concern the way in which people actually live through and interpret
situations. The bulk of the data collected in this PhD were indeed retrospective descriptions
(obtained through interview and to a limited degree through questionnaires).
Giorgi and Giorgi (2008) identify the following steps in phenomenological data analysis:
reading of the participant‟s entire description by the researcher;
constitution of the parts of the description: the text is divided into „meaning units‟;
transformation: meaning units are actively transformed by the researcher. For example,
what is implicit may be transformed to the explicit, allowing the analysis to reveal
meanings that are lived, but not necessarily clearly articulated by, or in the full
awareness of the participant; secondly, a degree of generalisation occurs; thirdly, an
attempt is made to describe what took place in a way which renders psychological
meanings more visible.
determining „structures‟: attempting to determine which constituents are typically
essential in order to account for the actual experiences reported, as well as the
relationship between the different constituents.
Interpretative Phenomenological Analysis
IPA is phenomenological, but with the added dimension of emphasising the active role of the
researcher in the research process. The researcher is attempting to get close to the
participant‟s personal world, but this access is complicated by the researcher's own conceptions;
these conceptions are in fact essential if the researcher is to make sense of that other personal
world through a process of interpretation. As previously explained, interpretation and reflexivity
were key aspects of the current analysis. As in phenomenology, participants‟ retrospective
accounts (often emerging from semi-structured interviews) usually constitute the main form of
data in IPA. Analysis is idiographic, starting with the detailed examination of the interview
transcripts, one at a time, before slowly working up to more general categorisations. Smith and
Osborn (2008) suggest the following steps in conducting an analysis:
Looking for themes in the first case: the researcher annotates the transcripts with
comments. Initial notes are transformed into concise phrases, and the themes move to a
slightly higher level of abstraction.
Connecting the themes: connections are then sought between the emergent themes, so
the order of themes becomes more analytical or theoretical.
Continuing the analysis with other cases, respecting both convergences and
divergences in the data.
Construction of superordinate themes across the different cases. By reaching higher
levels of abstraction, convergences may appear across seemingly disparate cases, so
the analysis respects both theoretical convergence and, within that, individual
idiosyncrasy in how that convergence is demonstrated.
It might be argued that these steps could be translated to any form of thematic qualitative
analysis, but the emphasis here is on understanding participants‟ meanings and acknowledging
the role of researcher interpretation, both key considerations in the current analysis. However,
as already discussed, the use of phenomenology or IPA in their pure forms was rejected
because of inconsistency with the epistemological approach of the thesis (there is no intention to
“bracket off” reality and study only the life world of participants).
Grounded theory
Finally, the present analysis was also informed by grounded theory, which emphasises the
generation of theory which is grounded in data. It assumes that social phenomena are complex
and therefore require “conceptually dense” theory, at various levels of generality, based on the
analysis of qualitative data. A grounded theory is inductively derived from the study of the
phenomena it represents: “one does not begin with a theory, then prove it. Rather, one begins
with an area of study and what is relevant to that area is allowed to emerge.” (Strauss and
Corbin, 1990, pp23). Grounded theory has been associated with the philosophical tradition of
symbolic interactionism (Snape and Spencer, 2003).
As the concepts and theory most relevant to this research were not clearly defined at the outset
in the project, grounded theory provided a useful approach to start to understand the role played
by informal information-sharing in travel behaviour. The use of grounded theory in transport
research featured only rarely in the literature review, although it should be noted that it is not a
research method in itself, but a framework in which to use and interpret findings using
appropriate methodology. Therefore, one might argue that this approach was used implicitly in
the design of some studies, particularly those using qualitative methods. As the present research
evolved it was felt appropriate to incorporate a certain degree of structure into the interview and
focus group schedules, which then informed the process of coding and analysis. Moreover, this
research did not emerge from vague concepts without any grounding in existing theory and
previous empirical work, as would have been required by a purist grounded theory approach.
The research approach was not therefore one of grounded theory in the strict sense, but more
akin to constructivist grounded theory, which acknowledges that researchers enter the research
scene not as impartial observers, but with an existing interpretive frame of reference (Charmaz,
This section has attempted to show how three specific approaches influenced the evolution of
thinking about the analysis within this thesis, but without being considered appropriate for
wholesale adoption.
Appendix E – List of NVivo codes, Phase 1
1. General travel behaviour
1.1. Reasons for modal choice to UWE
1.2. Reasons for modal choice for general travel
1.3. Choosing to live close to work or work close to where you live
1.4. Intentions to use other transport modes
2. Use of travel information
2.1. Understanding of the term travel information
2.2. In general, trips for which travel information is used
2.3. Travelling with other people
2.4. Visiting friends and family
2.5. Specific trips
2.5.1.Non-UWE trips - information considered
2.5.2.Information considered and/or used for first trips to UWE Formal travel information Informal travel information
3. Interaction between formal and informal information
3.1. Preference for formal or informal information
3.2. Combining formal and informal – complementarity
4. Attributes of informal information
4.1. Professional advice
4.2. Informal travel information
4.2.1.Specific contribution of informal information
4.2.2.Internet reviews or message boards
4.2.3.Difference between advice and information
4.2.4.Crops up in conversation
4.2.5.People who give me travel information
4.2.6.People tell you the bad things
4.2.7.Subconsciously absorbing information from others
5. Receiving travel information or advice from others
5.1. How it influenced my travel decision
5.2. Why it did not influence my decision
5.3. En route information
5.3.1.Choosing whom to approach
5.3.3.Being in the same situation
5.4. Pre-trip information
5.5. Attitudes to information content
5.6. Attitudes to or relationship with information giver
5.6.2.Having things in common (social identity)
5.6.3.Approval (subjective norm)
5.6.5.How well you know the person (incl. shared values)
5.6.6.Experience of information-giver Benefitting from other people's local knowledge Picking up on peoples' experiences
5.6.7.Reinforcing my own choices
5.6.8.Saving time or effort
5.6.9.Seeing that it was feasible
6. Making my own mind up
7. Giving information or advice to other people
7.1. Why other people give advice or information
7.2. Pre-trip advice to others
7.2.1.Influence of informal info on other people's decisions
7.2.2.Motivations Reciprocity Wanting to be directive Not wanting to impose your own views Making it easier for people social expectations Sharing your local knowledge
7.2.3.Thinking about other person's preferences
7.2.4.How well you know the person
7.2.5.Being confident of your knowledge
7.3. En route advice to others
7.3.1.Motivations empathy or reciprocity Making it easier for people social expectations being in the same situation Not wanting to impose own views how you are feeling Being confident of own knowledge
7.3.2.Not wanting to help people
8. Other factors influencing travel decisions
8.1. Own experience
8.2. Trial runs of time sensitive trips
8.3. Preferred travel habits
8.4. Giving it a try
8.5. Instrumental factors
8.6. Modal identities
8.7. Personal norms
8.8. Personality factors
8.8.1.Attitudes to uncertainty
8.8.3.Concerns about social norms of travel
8.8.5.Concern about lateness
8.8.6.Finding travel stressful
8.9. Cognitive factors
9. General interactions about transport
9.1. General conversations en route
9.2. General conversations when not travelling
10. Social norms
10.1. Norms of travel to and from UWE
10.2. Group norms
10.3. Injunctive social norms
11. Ideas about providing informal travel information at UWE
Appendix F – Post-Cycology Questionnaire
How would you describe your experience of using the Cycology website, and participating
in the project overall?
What was your impression of the information/comments posted on the website by the
other participants?
Were there any postings you particularly liked (or disliked) and why? Please give one or
two examples if you can.
Were there any postings you found particularly useful (or not useful) and why? Please
give one or two examples if you can.
How far did you feel the information was reliable, compared with, for example, routes
marked on a normal cycling map? Please explain your answer.
How did you feel about contributing information/posting questions yourself? (e.g
enthusiastic, reluctant, non-committal – please explain why)
Did you have any “off-line” conversations with any of the other participants, relating to the
project? If so, please expand (e.g. with whom, what about?).
Did you consider doing, or actually do, anything different (e.g. try a new cycle route) as a
result of reading comments on the website? Please explain your answer.
Did your participation in the project give you any other new ideas or thoughts relating to
cycling and travelling more generally? (if so, please explain)
What do you think would be the pros and cons of making interactive cycling information
such as this available for general use:
on a secure website within an organisation?
on a publicly accessible website?
Do you think this type of web-based information-sharing might work in relation to forms of
transport other than cycling? Please explain your answer.
Thank you for taking the time to fill in this questionnaire!
Please return to caroline.bartle@uwe.ac.uk and retain a copy for yourself.
Appendix G – Cycology interview guide template
Cycology Interview Guide template
Template to be modified for each interviewee, informed by observational and questionnaire data.
Interviews should preferably be held in a location with a computer and access to the website
(where this is not possible, printed screen shots will need to be taken to the interview).
Warm-up questions:
How would you describe your current level of cycling activity? (e.g. occasional, “fair
weather”, committed)
Turning to the Cycology project, can you describe how you used the website/email
digests? For example:
approximate frequency of reading and posting
prompted by email digest?
just read email digest without going to the site?
ignored digest?
read most postings, or only picked particular types or particular people (if so, follow this up
in main questions)
Main questions: social-psychological mechanisms
The following questions aim to reveal, inter alia, different types of trust (e.g. relational trust,
calculus trust) in the information and information-giver. There is some overlap in these
questions, so it is unlikely that every participant will be asked every question.
3. In your questionnaire, you said you liked this comment. Can you expand a little on why this
was? (if none was identified, ask them to browse website for a few moments to find an
example – encourage them to “think aloud” while doing so).
- content?
- use of language?
- something about the writer?
- reliability of the content/writer?
If participant does not talk about “usefulness” in their answer, ask them to also identify a
comment or set of comments they found “useful”. Ask them to expand on why it was useful,
using similar prompts.
4. Were there any comments you did not like? If so, what didn‟t you like about them?
Was there a participant (or a number of participants) whose contributions made a bigger
impression on you than others? (could add: for example, I can see from the data that you
looked at this person‟s postings a lot)
- why was this?
- what impression did you gain of that person?
- had you met or did you know anything about them beforehand?
- how reliable would you consider their postings to be, and why?
If interviewee is not forthcoming in providing examples, show him/her a few contrasting
examples of different types of posting (e.g. some very factual, some very chatty, some with
thumbnail photos, some using pseudonyms) and ask them to talk about them, using the same
prompts as above. .
Motivations for providing information to others (e.g. pro-social values, in-group
identity/”community sprit”, wishing to influence others):
5. How did you feel about posting comments yourself? (can you expand on what you said on
your questionnaire?/I note that you were an active contributor/I note that you only made a
few comments but you looked at other people‟s contributions…..)
For more frequent contributors:
- can you explain what motivated you to contribute?
- did the small group environment make any difference - would you have
made the same type of contributions to a public website?
- what did you think of the responses (if there were any) to your contributions?
For less frequent contributors:
- can you identify anything which prevented you from posting many comments? (e.g.
social inhibition, lack of time)
6. How would you feel (or did you feel) if/when someone said they were going to try your route
or otherwise follow a suggestion you had made? (show examples in cases where this had
actually happened)
7. How would you feel if someone who had not yet cycled to work saw advice/route
suggestions you had put on the website and gave it a go? (this is a theoretical question, as
there was no one in the project who was still only thinking about cycling to work).
Social identity
8. Did it matter that most participants worked at UWE like you?
OR: did it matter that most participants did not work in the same organisation as you? If yes,
would you have felt differently about the project if it had been internal to your own organisation?
9. Did the Cycology project help engender any sense of belonging to a “cycling community”? –
please expand.
Influence on self-reported attitudes, intentions and behaviour
10. Did your participation in the project lead you to do anything different (e.g. trying a new route)
/think about doing something different/ provoke any new thoughts about cycling or travel
more generally? (refer to questionnaire responses). Did you learn anything new?
11. If you acted on someone‟s advice within the project, or are thinking of doing so, can you say
why you considered their advice to be reliable? (could ask them to identify the relevant
posting and then explain what appealed to them about it – beyond the purely instrumental).
Wind down Questions
12. What is your overall assessment of the Cycology project?
- what worked well and did not work well?
- what would you improve?
- do you think it has the potential to encourage prospective and hesitant cyclists?
13. Do you think anything could have improved the social interaction?
14. What suggestions might you make to a website designer?
15. How do you think this sort of informal information system compares with online, door-to-door
route planners? (could show an example)
16. In what kind of context could online information-sharing such as this best be applied? E.g.
privately within an organisation/on a public website/in relation to transport other than
cycling? (refer to questionnaire answers)
17. Do you think any of the information collected in the Cycology project should be kept? (e.g.
routes imported onto public Bike layer of BristolStreets website). If so, need to discuss
details, as each route/comment will only be made public with the consent of the person who
originated it.
18. Do you have anything else you would like to add?
Thank you for your participation!
Appendix H – Cycology Guide for Participants
Caroline Bartle
Research student
Centre for Transport & Society, UWE
0117 32 81430
Cycology Project:
Cycling and on-line information-sharing
Participant Guide
What the project involves
Thank you for agreeing to take part in this research project about the sharing of local cycling
information, using a secure section of www.bristolstreets.co.uk, a local map-based website
providing transport and other local information. You are one of a small group (about 25 people)
who have been recruited to the project to share cycling routes and other information with one
another over a period of 6 weeks, starting on Monday 8 June. The group comprises people with
different levels of cycling experience who are currently cycling to work or study (or contemplating
doing so) at a number of neighbouring organisations in North Bristol.
Although the project runs for 6 weeks, it should only take a few minutes of your time every day
or so, at a time convenient for you. You could mark your favourite cycling route/s on the
interactive map, post comments or photographs, and ask questions about local cycling matters.
You will receive a regular email digest of new postings on the website. At the end of the six
weeks you will be invited to a face-to-face interview to discuss your experiences of taking part in
the research.
Using the Cycology layer of the website
Setting up an account on www.bristolstreets.co.uk
Because this area of the website will be visible only to the project participants, you will
need to sign in with a password in order to access it. You will need to register on the site
to create your log in password associated with your email address. If you intend to use an
email address other than the one you have already given as your contact email, please let
me know this other address in advance (preferably by email) as access to the research
layer will only be given to email addresses which are known to be participants.
To register, open the website home page, and click on Sign in in the top right hand
corner. The email address you use will be the one to which your email digest will be sent.
Make up a password and click the sign in.
You will then receive an email and will need to follow the hyperlink to confirm your
registration. Once this is confirmed, you log in using the same sign-in box you used to
Once logged-in, click on the home icon on the side menu to make the cycology icon
appear. Click on the cycology icon to enter the project area.
Choosing your account settings
Once signed in, click on the my stuff icon, followed by account icon, on the side menu.
Set your “how may we use your email to contact you?” to announcements (this is the
email digest).
Set your site name (this will appear on all your postings). If you do not change this it
defaults to the part of your email address before the “@”. You can use your real name or
a nickname if you prefer.
Please click the “Add/Change Picture” button to upload a picture to appear as a thumbnail
with your postings. The thumbnail will help other users recognise your posts. It does not
necessarily have to be a portrait, but that is what most people will use11.
Click on home to return to the bristolstreets home page.
Click on cycology to return to the project area.
Switching between cycology and the rest of the bristolstreets website
You can switch to other parts of the website without logging out by clicking the home icon.
To return to the project area, simply click on the cycology icon. Please check that you are
in the project area before posting your comments, routes etc.
A word of warning: if you click on the bike icon on the bristolstreets home page, you will
enter a different layer of the site where you will see markers posted by cyclists as part of
the City Council cycling survey. Any comments you add here will be publicly visible as this
is not part of the cycology project. If in any doubt, check the side menu - if you see the
cycology icon in the top right corner of the menu, you are in the right place!
Using different map views
You can look at the map in either terrain mode or satellite mode by clicking the mountain
or satellite buttons on the tool bar along the top. You can choose to have road names and
colours denoting main roads visible by clicking the road sign button between the terrain
and satellite buttons.
Please note that that all information uploaded by participants to the cycology area of the
website (names, photos, travel information etc) will only be visible to other participants, the
researcher, and the website administrator. You will not be identified in the reporting of the
research - even if you choose to use your real name and photo within the project, these will be
anonymised in the writing-up of the research.
Choosing the types of information you want to see
You can turn different types of information on and off by checking the boxes on the side
menu when you are on the home page of the cycology area:
a. Existing overlays and markers
Cycle paths and cycle lanes
Quiet and recommended roads
Cycle parking
Cycle shops
b. Markers created in the project
Visitor markers
Only show markers with routes
Show comment titles on map
The visitor markers have different colours and shapes as each identities a
different participant.
Looking at routes and creating new ones
To see a route created by another participant, you need to click on the relevant marker,
and the green line marking the route will appear. Only one green line is visible at a time.
To be able to get a quick idea of which routes you might want to look at, check the only
show markers with routes box, and the show comments titles on map box.
You can also choose which route to look at by clicking on the comments icon and
checking the comments with routes only box. Clicking on a route comment on the side
bar will open up the marker on the map.
To add a new route, click on the add route icon on the side menu. Instructions for drawing
lines on the map can be found on side bar. When you have finished drawing the line, click
save route, and a comments box will appear in which you can write the route origin and
destination and a route description (and upload a photograph if you like). When you save
this, the marker will appear on the map at the beginning of the route. If you want to add a
comment or photo along a route, you will need to create the route in separate segments,
or add an additional comments marker later.
Reading and adding comments
If you want to add a new comment to the map, simply move your mouse to the position
you would like to put it, and click once quickly.
NB it is quite easy to create a new marker in this way by accident – if this happens, just
click the cancel button when the comments box opens.
You can also create a non-geographical comment by clicking the start a floating
comment button.
The quickest way to read and respond to an existing comment is by clicking on a visitor
marker on the map, which opens the comment box, you can then click the “add comment”
link in the lower part of the comment box.
Clicking the comments icon on the cycology home page allows you to read comments on
the side bar and filter them. You can choose to:
Show comments with routes only
Show floating comments only
Show comments by individual participants
Clicking on a comment on the side bar will open up the marker on the map.
The email digest
You will receive a regular (a maximum of once a day) email digest containing all new
postings on the cycology layer of the website. Clicking on a hyperlink will take you directly
to the new comment on the webpage.
Further help
There is a help feature at the bottom of the screen on bristolstreets, but if you have any further
questions about using the website, or any other aspects of the project, please do not hesitate to
contact me.
Confidentiality, data protection and code of conduct
Your attention is drawn to confidentiality and data protection section of the project information
sheet, and to the code of conduct for on-line discussion. In line with UWE research ethics
requirements, you will also be asked to sign a consent form before starting the project.
And finally……
I hope you will find participating in this project to be an enjoyable and useful experience. As well
as helping me with my PhD research, I hope that it might make a contribution (however small!)
to building a “cycling culture” in this part of Bristol, as well as helping to inform discussion on the
value of web-based “peer-to peer” travel information.
Caroline Bartle, 27/05/09
Appendix I – Information sheet for Cycology participants
Cycology Project:
Cycling and on-line information-sharing
Project Information Sheet:
This project constitutes a specific case-study within a 3-year PhD study exploring word-of-mouth
influences on everyday travel behaviour .
What is the research about?
The aim of this research is to explore information-sharing, via the internet, amongst a small
group of people (approximately 25) who are either already cycling, or considering cycling, to a
number of neighbouring organisations in North Bristol.
What will it involve?
You will be invited to share cycling routes and other information with one another using a
specially designed, secure section of the website www.bristolstreets.co.uk. The research runs
over a period of 6 weeks, starting on 8 June 2009, but should only take a few minutes of your
time every day or so (whenever convenient for you). You could mark your favourite cycling
route/s on the interactive map, post comments or photographs, and ask questions about local
cycling matters. At the end of the six weeks you will be invited to a face-to-face interview to
discuss your experiences of taking part in the research and any influence you think it might have
on your commuting behaviour. I may also contact you occasionally by email or telephone during
the course of the experiment.
Full title: Exploring word-of-mouth influences on travel behaviour: the role of informal
information and its implications for advanced traveller information systems.
Confidentiality and data protection
Content added to the secure area of the website will be visible only to the research participants,
the PhD researcher and the website administrator during the experiment. By default it will be
removed from the website and anonymised on completion of the experiment. However, you may
request any content you have created to be transferred to the main (public) area of
www.bristolstreets.co.uk after the experiment (for example if you think the cycle routes you have
marked would be of interest general users of the website). Your browsing activity on the website
will be recorded during the course of the project in order to provide basic quantitative data, such
as the most and least frequently looked at routes and comments on the website, but this will
cease immediately upon completion of the project. Any such data will be anonymised and
securely stored.
Your name will not appear in any reporting of the research results: all participants will be given a
pseudonym and participant anonymity will be maintained in my PhD thesis and any related
academic publications and presentations. All information will be kept securely and only shared
within my PhD supervision team. Your participation is voluntary, and you are free to withdraw
from the study if you wish at any time.
On-line discussion “code of conduct”
Participants are requested to read the project code of conduct for on-line discussion, based on
standard terms of participation in on-line forums.
PhD Student:
Director of Studies:
Caroline Bartle
Dr Erel Avineri
Dr Kiron Chatterjee
Tel: 0117 32 81430
Centre for Transport & Society, School of the Built and Natural Environment, Faculty of
Environment and Technology, University of the West of England, Bristol.
Appendix J – Code of conduct for on-line discussions
Cycology Project:
Cycling and on-line information-sharing
Code of conduct for on-line discussions
In line with standard rules of participation in on-line discussion forums, it is assumed that
you have read, understood and will abide by the following :
About your posts
Keep your contributions civil and relevant. We're committed to providing an atmosphere in which
friendly and mature dialogue takes place.
The following are not acceptable:
Messages which are unlawful, harassing, defamatory, abusive, threatening, harmful, racially
offensive, homophobic, or are highly objectionable.
Impersonating someone else or attempting to mislead other members of the forum about your
Stay on Topic - Try to keep posts on the subject of the thread.
Although you might choose to use your real name and a thumbnail portrait on the website, do
not reveal any other personal information about yourself (for example: your telephone number,
home address etc).
Respect the confidentiality of other participants in the project – they may not wish their postings
to be repeated outside the project group.
Based on UWE students‟ union discussion forum rules
Keep graphics that you use in footers and avatars etc to reasonable dimensions.
You agree that the webmaster, administrator and moderators of this forum have the right to
remove, edit, move or close any topic/content at any time should they see fit.
About the law
You may not post any defamatory or illegal material of any nature on the project website. This
includes text, graphics, video, programs or audio. Posting a message with the intention of
committing an illegal act is strictly prohibited.
You agree to only post materials to which you have the copyright or other permission to
distribute electronically. You may not violate, plagiarise, or infringe on the rights of third parties
including copyright, trademark, trade secret, privacy, personal, publicity, or proprietary rights.
Please take a moment to check the terms of use, including the privacy policy, of
Please note however, that your browsing activity on the website will be automatically recorded
during the course of the project in order to provide the researcher with basic quantitative data,
such as the most and least frequently looked at routes and comments on the website. This will
cease immediately upon completion of the project. Any such data will be anonymised and
securely stored.
Appendix K – Categories of website post over time
The charts below show how the three groups of message categories depicted in Figure 8-4 were
distributed over the six weeks of the project. Figure K 1 shows that all route-related posts
peaked in the first week.
Figure K 1: Numbers of route-related posts over time
Numbers of route-related posts over time
Route with comment
Further detail of own
Response to another
person's route
15-21 22-28 29 June 6 - 12 13 -17
June June - 5 July July
The second group of categories, shown in Figure K 2, comprise direct questions and responses
to them; many of the questions concerned routes (and therefore are not entirely distinct from the
“route-related” categories), but they also covered a broader range of topics, such as taking
bicycles on trains or the Bristol Cycling City scheme. There were slightly more questions than
answers, and a slight time lag between the two, with most questions in the first and third week
and most responses in the fifth week.
Figure K 2: Questions and responses over time
Questions and responses over time
Responses to
22-28 29 June 6 - 12
June - 5 July July
13 -17
The third group of categories (general comments and responses), shown in Figure K 3,
comprised proactive comments concerning non-route related matters, plus a small number (5) of
responses to them. These included comments about the behaviour of other road users,
warnings (such as broken glass), notification of an event, and sharing experiences such as
seeing deer along the cycle path, or the friendliness of cyclists in the rain. The pattern of these
postings over time is shown in . The discrepancy in numbers between these types of proactive
comment and the responses to them may be explained by the fact that many did not directly
invite a response. Alternatively, this may be an indication that the website was not seen primarily
as a forum for general social exchange, but rather as a place to share specific information and
ask and answer specific questions.
Figure K 3: General comments and responses over time
General comments and responses over time
General comments
8-14 15-21 22-28 29 6 - 12 13 -17
June June June June - July
5 July
Responses to
general comments
Appendix L – List of NVivo codes, Phase 2
Anonymity issues
Behavioural and attitudinal
effects of Cycology
Complaining or lobbying for change
Heeding warnings
Looking for things to tell people
Looking out for features which others have
Real world contacts
Reasons for no effects
Reinforcing existing attitudes and
Showed Bristolstreets to others
Thinking of others cycling the same route
Trying new routes
Comparison between cycling
and other transport info
Comparison with cycling
discussion forums
Comparisons with other types
of map or planner
Comparing map with user comments with
a normal map
Comparing map with user comments with
journey planners
Google maps
Consequences of this being a
research project
How it did not change things
How it changed things
Cycling culture within the
Cycology as encouragement
for cycling
Practical information
Social support
Too many other barriers
Would have been useful when starting or
going somewhere new
Effects of group size and
Effects of group size
Within workplace or involving others
Evaluation of Cycology
Campaigning potential
Comments on social interaction
Effects of the focus on commuting
Getting something out of it (or not)
How project could have been improved
Informative, useful (or not)
Liked or enjoyed it
Not being a sociable person
OK or negative
10. Evaluation of specific
Advice on practical matters
Community building and support
Creating a particular image of the route
Disapproving, patronising or rude
Helpful, useful (or not)
No particular impression
Sharing a good or bad experience
11. Evaluation of techncial
Overall evaluation
Suggested improvements
Technology as a draw of the project
Things which did not work well
Things which worked well
12. FTF interactions about
13. Getting an impression of
other participants
Assuming postings are written by men
perceptions of
Seeing more men
than women cycling
Cycling activists or experts
Difference between cycling in town and
Different cycling styles
Different levels of confidence, experience
and fitness
Hard to get any impression
Ordinary people
Use of portraits or icons
Use of pseudonyms
Working out who they are
14. How the Cycology system
was used
Digest and website
For advising, getting advice or both
Looking while at work
Only reading what was geographically
Relying on the email digest
15. Participants' general cycling
activity and attitudes
cycling as local transport
cycling for leisure
Particpants' reasons for cycling (as
Because it is quicker
or take a guaranteed
Cycling makes life
Health and wellbeing
perceptions of safety
Wanting to encourage others to cycle
16. Perceptions of different
types of cyclist
Friendly, helpful
17. Reasons for contributing to
Encouraging people
Give and take
Just to make a contribution to the project
Knowing another participant
Nice to help people
Seeking moral support
Taking the glory
Wanting to share a gripe
Wanting to share pleasant things
18. Reasons for not
contributing to website
Feeling criticised
Lack of time
Limited knowledge
Not wanting colleagues to think they are
wasting time
Not wishing to give inappropriate advice
Preferring to read rather than write
Technically awkward
19. Reliability of information
Assuming everyone's intentions are good
Comparison with credibility of an open site
Comparison with own experience
Level of detail
Perceived attributes, views, experience of
Reliable but no explanation
User info is more reliable than official info
20. Sense of community
Communities of other transport users
Community of cyclists generally
Community of cyclists within the project
Community of people who had volunteered
for the project
Compared or contrasted with open forums
Feeling excluded from project community
People with similar cycling styles to me
Work-based community
Appendix M – Titles of posts and threads added to the
Cycology website
Chronological as they appeared on the website, coded by theme:
Wayfaring (route information)
Safety (vehicles or personal security)
Sharing space with other road/path users
Sensory/aesthetic/affective descriptions
Other (includes events, warnings, wider discussion points)
Clifton Village to Dupont Building UWE
Little Stoke to UWE
Redland to UWE
Holly Bush Lane (Stoke Gifford to UWE (Frenchay)
Building works
YMCA cricket ground
Not a cycle path
Fishponds to UWE
Quiet bridge over railway
Footpath/cycle path
Winterbourne Road Crossing
Three Brooks Nature Reserve
Brook Way
Staple Hill to Frenchay
Gloucester Rd (bottom) to Clifton with minimal hills
Horfield to Stoke Gifford
Stapleton Road quieter?
Whitehall to UWE Alternative Fishponds to UWE route
Alternative Fishponds to UWE route
Staple Hill to UWE, route 2
Cycle shops
Horfield to UWE Frenchay campus
Gainsborough Square
View across Bristol
Dovercourt Road
Staple Hill to UWE
Montpelier to UWE
Through the park
Kingswood to UWE
Bromley heath to HP/UWE
Hambrook to HP/UWE
Arches to Clifton village
HBishopston to Bristol/Bath railway path
Cycle days out
East-West access
Staple Hill to Keynsham
Big Bike Breakfast
Cycling in the rain
Traffic tie-up question
Redland to UWE
Horfield to HP/MoD/UWE
Muller Road to UWE
Downend to HEFCE
St Andrews to UWE
St Andrews to UWE (via Lockleaze Hill)
St Andrews to UWE (short n‟ nasty)
BS32 to BS16 (UWE)
Scenic view
Cycle path
Traffic island is too narrow
Cycle lane not very helpful
Bishopston to UWE
UWE to Bishopston
Eighth Avenue – can you get through here yet?
Gripe with pedestrians
Broken glass on this small roundabout this morning
Mangotsfield station – fun for young kids
Montpelier to Temple Meads
Bikes on a train
BRI – Lime Kilm Close, Filton
Cycling for Leisure
Stoke Bishop to Temple Meads
Less traffic alternative to (most of) Whiteladies Road
Adding photos?
Deadly roundabout
Any info on this area?
Bristol Cycling City Award
Oh deer!
Route when feeding neighbour‟s cat
Road resurfacing
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF