User manual | ISM research

THE EXTENT, NATURE AND EFFECTS
OF FOOD PROMOTION TO CHILDREN:
A REVIEW OF THE EVIDENCE TO
DECEMBER 2008
PREPARED FOR THE WORLD HEALTH ORGANIZATION
By Georgina Cairns, Kathryn Angus & Gerard Hastings
Institute for Social Marketing, University of Stirling &
The Open University, United Kingdom
December 2009
WHO Library Cataloguing-in-Publication Data:
The extent, nature and effects of food promotion to children: a review of the evidence to
December 2008 / by Georgina Cairns, Kathryn Angus and Gerard Hastings.
1.Food. 2.Marketing. 3.Child. 4.Advertising. 5.Food supply. 6.Diet - trends. I.Cairns,
Georgina. II.Angus, Kathryn. III.Hastings, Gerard. IV.World Health Organization.
ISBN 978 92 4 159883 5
(NLM classification: QT 235)
© World Health Organization 2009
All rights reserved. Publications of the World Health Organization can be obtained from WHO Press,
World Health Organization, 20 Avenue Appia, 1211 Geneva 27, Switzerland (tel.: +41 22 791 3264; fax:
+41 22 791 4857; e-mail: [email protected]). Requests for permission to reproduce or translate
WHO publications – whether for sale or for noncommercial distribution – should be addressed to WHO
Press, at the above address (fax: +41 22 791 4806; e-mail: [email protected]).
The designations employed and the presentation of the material in this publication do not imply the
expression of any opinion whatsoever on the part of the World Health Organization concerning the legal
status of any country, territory, city or area or of its authorities, or concerning the delimitation of its
frontiers or boundaries. Dotted lines on maps represent approximate border lines for which there may
not yet be full agreement.
The mention of specific companies or of certain manufacturers’ products does not imply that they are
endorsed or recommended by the World Health Organization in preference to others of a similar nature
that are not mentioned. Errors and omissions excepted, the names of proprietary products are
distinguished by initial capital letters.
All reasonable precautions have been taken by the World Health Organization to verify the information
contained in this publication. However, the published material is being distributed without warranty of
any kind, either expressed or implied. The responsibility for the interpretation and use of the material
lies with the reader. In no event shall the World Health Organization be liable for damages arising from
its use.
The named authors alone are responsible for the views expressed in this publication.
ACKNOWLEDGEMENTS
Our thanks to Dr Louise Hassan for rating studies for quality and potential causality, to Dr Ingrid
Holme, Ray Kent and Gill Cowburn for their input and to Aileen Paton for administrative support. Our
thanks also to the authors of the previous Institute for Social Marketing reviews on food promotion to
children on whose original work this update is based.
CONTENTS
EXECUTIVE SUMMARY
1
BACKGROUND
1
REVIEW METHODS
1
EXTENT AND NATURE OF FOOD PROMOTION TO CHILDREN
1
THE EFFECTS OF FOOD PROMOTION TO CHILDREN
2
1. INTRODUCTION
4
NOTES AND COMMENTS ON THE 2008 UPDATE
4
2. REVIEW QUESTIONS
5
EXTENT AND NATURE OF FOOD PROMOTION TO CHILDREN
5
EFFECTS OF FOOD PROMOTION TO CHILDREN
5
3. REVIEW METHODS
6
3.1 THE SYSTEMATIC REVIEW PROCESS
6
3.2 THE SEARCH STRATEGY
6
3.2.1 PREVIOUS SYSTEMATIC REVIEWS
6
3.2.2 UPDATE OF DATABASE SEARCHES
6
3.3 INITIAL RELEVANCE ASSESSMENT
9
3.4 ASSESSMENT OF STUDY QUALITY
10
3.5 OTHER EVIDENCE ON THE EXTENT AND NATURE OF FOOD PROMOTION IN LOW AND MIDDLE
INCOME COUNTRIES
11
4. THE EVIDENCE BASE
12
4.1 THE EXTENT AND NATURE OF FOOD PROMOTION TO CHILDREN
12
4.1.1 WHAT PROMOTIONAL CHANNELS ARE USED BY FOOD MARKETERS TO REACH
CHILDREN?
12
4.1.2 WHAT FOOD ITEMS ARE PROMOTED TO CHILDREN?
14
4.1.3 WHAT CREATIVE STRATEGIES ARE USED BY FOOD MARKETERS TO TARGET
CHILDREN?
16
4.1.4 CONCLUSION
18
4.2 FOOD PROMOTION AND MARKETING IN DEVELOPING AND MIDDLE INCOME COUNTRIES
18
4.2.1 WHAT MARKETING STRATEGIES ARE BEING USED TO PROMOTE FOODS IN
DEVELOPING AND MIDDLE INCOME COUNTRIES?
18
4.2.2 CONCLUSION
23
4.3 THE EFFECTS OF FOOD PROMOTION TO CHILDREN
24
4.3.1 HOW DO CHILDREN RESPOND TO FOOD PROMOTION?
24
4.3.2 IS THERE A CAUSAL LINK BETWEEN FOOD PROMOTION AND CHILDREN'S FOODRELATED KNOWLEDGE, ATTITUDES, BEHAVIOURS AND HEALTH STATUS?
27
4.3.3 CONCLUSION
32
4.4 WHAT IS THE EXTENT OF FOOD PROMOTION'S INFLUENCE RELATIVE TO OTHER FACTORS? 33
4.5 DOES FOOD PROMOTION AFFECT TOTAL CATEGORY SALES, BRAND SWITCHING OR BOTH? 34
5. DISCUSSION
36
REFERENCES
37
EXTENT AND NATURE REFERENCES
37
EFFECTS REFERENCES
44
OTHER REFERENCES CITED
48
ANNEX 1:
RECORD OF SEARCHES
ANNEX 2:
DATA EXTRACTION TABLE: THE EXTENT AND
NATURE OF FOOD PROMOTION TO CHILDREN
ANNEX 3:
A1-A4
A5-A63
DATA EXTRACTION TABLE: THE EFFECTS
OF FOOD PROMOTION TO CHILDREN
A64A117
EXECUTIVE SUMMARY
Background
This document reviews evidence to December 2008 on the
global extent and nature of food promotion to children, and its
effects on their food knowledge, preferences, behaviour and dietrelated health outcomes. The review was commissioned by the
World Health Organization (WHO) and updates a systematic
review of the evidence conducted on behalf of WHO in 20061.
The 2006 review in turn built on earlier reviews commissioned by
2
3
the UK’s Food Standards Agency and WHO . All these reviews
were conducted by the Institute for Social Marketing (ISM) at the
University of Stirling, and the Open University, United Kingdom
of Great Britain and Northern Ireland.
Review methods
Systematic literature review methods were used to identify peer
reviewed published literature via systematic searches of
academic literature databases. Additional searches were
conducted online to identify ‘grey’ literature. The titles and
abstracts yielded by the searches were assessed according to
pre-defined relevance criteria for the review. This process
identified 115 studies on the extent and nature of food promotion
to children and 90 on its effects. The results of the literature
search are summarized in the evidence tables in Annexes 2 and
3.
Extent and
nature of food
promotion to
children
Studies examining the extent and nature of food promotion to
children consistently conclude that food promotion is the most
prevalent marketing category targeting children and young
people. Content analysis research finds that the majority of foods
and food products promoted are energy dense, high fat, sugar
and/or high salt, and in sharp contrast to national and
international dietary guidelines. Sugar-sweetened breakfast
cereals, soft-drinks, confectionary and savoury snacks are the
most frequently advertised categories, with fast-food promotion
continuing to gain marketing share. Promotion of unprocessed
foods, such as fruit and vegetables, wholegrain and milk is found
to be almost zero.
The themes used to promote child-oriented food products to
children focus on fun, fantasy, novelty, plus taste. In contrast,
parent-targeted marketing of the same child-oriented product
emphasizes nutrition and health themes.
Early research focuses on food promotion through television (TV)
advertising. More recent research has examined other
promotional channels and strategies. Limited trend analysis data
suggests that there is a shift of marketing spend from TV
1
Hastings G et al. (2006). The Extent, Nature and Effects of Food Promotion to Children: A Review of the
Evidence. Geneva, World Health Organization.
2
Hastings G et al. (2003). Review of Research on the Effects of Food Promotion to Children. Glasgow,
University of Strathclyde, Centre for Social Marketing.
3
McDermott L et al. (2004). Desk Research to Examine the Effects of Food Marketing on Children. Glasgow,
University of Strathclyde, Centre for Social Marketing.
1
advertising to other forms of promotion, although TV advertising
remains the most dominant medium to date. As well as growth in
promotion through emerging promotional platforms, such as the
Internet, viral marketing by SMS (short messaging service) and
more interactive competitions, games and membership, there is
evidence of increasingly sophisticated integrated marketing
strategies. Some of the research included in the review notes
that the multiple channels and sources complement, and direct
activity to, one another. As well as the multiplying benefits of
integrated marketing activities, this type of strategy can be used
to establish and maintain brand loyalty with multiple relationshipbuilding benefits and reward offers.
The research also recognizes that in addition to explicit childtargeted marketing, children are exposed to a great deal of food
promotion through generic medium such as mainstream TV, onpack promotions and sponsorships. The evidence base for low
and middle income countries is less substantial than for
developed economies. The research available to date however
does indicate that marketing strategies in lower income countries
follow very similar patterns to more affluent markets, with the
emphasis on low nutritional quality foods, the dominance of TV
advertising but increasingly supplemented with other forms of
promotion that reinforce marketing objectives and build brand
awareness and allegiance.
The evidence also indicates that marketing to children
approaches them both as consumers in their own right and as an
access point to wider markets. A number of studies report that
children have independent spending power and significant
influence over household and parental purchasing decisions. In
addition, some of the research reviewed highlights the role
children play in developing
new market opportunities,
particularly in low and middle income countries where new
market opportunities are being developed. Fast-food and quick
service restaurants case study research suggest that childrenfriendly initiatives have been effective entry strategies in fast
growing markets in low and middle income countries.
The effects of
food promotion
to children
Survey evidence confirms that many forms of food product
promotions are popular with, and engage children. On-pack
promotions, advertising, free gifts, and many other marketing
techniques encourage interest and purchase. Research on recall
of food advertisements in particular finds very high levels of
awareness and enjoyment. Parents report, and are observed to,
frequently accede to purchase requests. There is also some
evidence that socially disadvantaged mothers attribute more
credibility to food adverts than more privileged peers.
Evidence from more complex studies, assessed as capable of
inferring causality, find promotional activity is having an effect on
children. Research that has compared the influence of other
factors on food behaviours, determinants of behaviour and health
outcomes, find that family and parents, peers, other lifestyle
factors, as well as socioeconomic status are also important.
Nevertheless, the research to date suggest that food promotion
has an equally important effect.
The evidence reviewed indicates that food promotion does have
2
a modest impact on nutrition knowledge, food preferences and
consumption patterns. A limited number of studies that have
examined exposure to food promotion and diet-related health
status such as adiposity find positive associations between
exposure to food promotion and adverse health outcomes. The
evidence that food promotion influences purchasing behaviour is
of modest strength. On balance, the evidence also indicates that
the influence extends to category as well as brand level
awareness and choices.
In summary, the review confirms that in both developed and
developing countries: there is a great deal of food promotion to
children. Television advertising is the most dominant promotional
channel but the full range of promotion and marketing techniques
and strategies are used in, and integrated together, by the food
and advertising industries. Children recall, enjoy and engage with
the multiple promotions and evocative brand building initiatives
they are exposed to. The emergence of new mass media
channels such as website and mobile telephone SMS services
offer less visible but highly direct targeted marketing
opportunities. The evidence base for the effect and reach of
these newer promotional channels is quite small, but to date,
suggests it is gaining share rapidly and effectively.
The evidence reviewed confirms that the food products promoted
continue to represent a very undesirable dietary profile, with
heavy emphasis on energy dense, high fat, high salt and high
sugar foods, and almost no promotion of foods that public health
evidence encourages greater consumption of – for example fruit
and vegetables.
Research that has examined associations between food
promotion and food behaviours, determinants of behaviour and
diet-related health outcomes, finds modest but consistent
evidence that the link is causal. This evidence base is drawn
from research conducted in the developed world. It is however,
reasonable to assume these findings can be extrapolated to low
and middle income countries because the research on extent
and nature of promotion to children confirms that marketing
activity in poorer countries mirrors the activity in more affluent
economies. Furthermore, there is evidence that children in
emerging markets are targeted both directly as consumers and
as a bridgehead to wider society experiencing rapid cultural
change. There is also a possibility that with a shorter history of
exposure to sophisticated marketing techniques and less
regulated marketing controls, children in developing and
emerging economies may be more susceptible to the persuasive
influences of food marketing and promotions.
3
1. INTRODUCTION
The report updates three previous systematic reviews of the
nature, extent and effects of food promotion to children (Hastings
et al., 2003; McDermott et al., 2004; Hastings et al., 2006). The
Hastings et al., 2003 review, undertaken for the United Kingdom
Food Standards Agency (FSA), was the first ever systematic
study of the effects of food promotion on children. The
McDermott et al. 2004 systematic review was commissioned by
the World Health Organization (WHO) to broaden the evidence
base and examine studies on the effects of food marketing to
children in developing countries. Finally, the Hastings et al. 2006
review was produced for WHO as a Technical Paper in
preparation for the WHO forum and technical meeting of
stakeholders on ‘Marketing food and non-alcoholic beverages to
children’, held in Oslo (May 2006). Since its publication, a
number of other reports and review papers on food marketing to
children and young people have been published, including a very
comprehensive study by the United States of America Institute of
Medicine (2006). The growing body of evidence is in broad
agreement that food and beverage marketing to children is
extensive and does have measurable effects on diet and dietrelated outcomes.
Notes and
comments on the
2008 update
This document summarizes a review of the extent and nature of
food promotion to children, and its effects on their food
knowledge, preferences and behaviour for the period 2006 to
2008. The research was undertaken by the Institute for Social
Marketing at the University of Stirling and the Open University,
the United Kingdom of Great Britain and Northern Ireland, on
behalf of WHO in preparation for the WHO Ad-Hoc Expert Group
Meeting on ‘Marketing of foods and non-alcoholic beverages to
children’ held in Geneva (1-4 December 2008).
The same systematic search strategies were used for this update
as for the 2006 review undertaken for WHO. Research from low
and middle income countries was located through these search
strategies. However, a separate search for extent and nature
studies from low and middle income countries was not
conducted.
4
2. REVIEW QUESTIONS
The review aimed to examine the extent and nature of food
promotion to children and its effects on children. The following
research questions were addressed:
Extent and
nature of food
promotion to
children
Effects of food
promotion to
children
1. What promotional channels are used by food marketers
to reach children?
2. What food items are promoted to children?
3. What creative strategies are used by food marketers to
target children?
1. How do children respond to food promotion?
2. Is there a causal link between food promotion and
children’s food knowledge, preferences and behaviour?
3. If food promotion is shown to have an effect on
children’s food knowledge, preferences and behaviour,
what is the extent of this influence relative to other
factors?
4. In the studies which demonstrate an effect of food
promotion on children’s food knowledge, preferences
and behaviour, does this affect total category sales,
brand switching or both?
5
3. REVIEW METHODS
This section describes the methods used for the review.
3.1 The
systematic
review process
Systematic review methods were used to search for, identify, and
assess evidence for the review. A systematic review is ‘a review
of the evidence on a clearly formulated question that uses
systematic and explicit methods to identify, select and critically
appraise relevant primary research’ (Khan et al., 2001). The
process involved a number of key stages (see Figure 1).
3.2 The search
strategy
As shown in Figure 1, the first task was to develop suitable
research questions for the review. These questions are outlined
in Section 2. The search strategy for the review was then
developed. As this review is an update to a previous review
(Hastings et al., 2006, referred to as the “2006 Review” in this
Section), adding the evidence published between March 2006
and mid-November 2008, the methodologies to this current
review and the previous version are both outlined. Two main
approaches were used to search for relevant research, as briefly
described below.
3.2.1 PREVIOUS
SYSTEMATIC
REVIEWS
Two previous systematic reviews provided the starting point for
the 2006 Review. The bibliographies of the review of food
promotion to children (Hastings et al., 2003) and an additional
review focusing on food promotion to children in developing
countries (McDermott et al., 2004) were systematically
examined. As these earlier reviews had very similar remits to the
current review, and applied similar inclusion criteria, all of the
studies cited in these earlier reviews were included in the 2006
Review.
In turn, this review includes all the studies included in the 2006
Review.
3.2.2 UPDATE OF
DATABASE
SEARCHES
New searches were undertaken in November 2008 to identify
relevant studies published since March 2006 (the cut-off date for
the 2006 Review’s literature searches). Two main methods were
used: searches of academic databases and searches for ‘grey’
literature (i.e. unpublished papers or reports of sufficient quality
to be reliable and verifiable) via online databases and websites.
(i) Academic
databases
The primary source for literature was academic databases.
Systematic searches were undertaken on the 15th November
2008 for the date range 2006 to 2008 in seven academic
database interfaces (covering 11 different academic literature
databases). Search strategies for the two original reviews
6
(Hastings et al., 2003 and McDermott et al., 2004) were
combined to run more efficient searches and minimize duplicate
records. Details of every search were documented fully to
provide a transparent and replicable record of the review
process. Full details of the strategies used are in Annex 1.
This 2008 search strategy has excluded the IngentaConnect and
ScienceDirect databases as there was a great deal of overlap
between all the academic literature databases and these appear
to be publishers’ databases rather than subject databases. The
database Business Source Elite was added, as its peer-reviewed
content is relevant to the topic, containing business,
management and economics literature. An extra Medline search
using a different interface was also added.
Citations from all the searches were downloaded into
ReferenceManager® software for storage and duplicates were
deleted. The sample of papers to be assessed against the
relevance criteria was 728.
7
Figure 1: Overview of review process
Development of review questions
Development of search strategy and
initial relevance criteria
Searches for relevant literature
• Based on previous review, Hastings et al.
2006, which used reference lists and search
strategies from two previous systematic
reviews (Hastings et al., 2003; McDermott et
al., 2004)
• Update searches from previous systematic
review to identify new studies via:
• electronic academic databases
• electronic grey literature searches
Initial relevance criteria (see
Box 1 below)
Initial relevance assessment
Does not meet
criteria
Meets criteria
Exclude
Retrieval of full text of relevant
studies
Assessment of study quality
Data synthesis and analysis
REPORT
8
(ii) Grey literature
A number of online databases and websites previously searched
for the 2006 Review were searched for the date range 2006 to
rd
2008 on the 23 March 2009, including the Advertising
Education Forum database of research on advertising and
children; ESRC Society Today, an online electronic database of
social science research funded by the Economic and Social
Research Council; WHOLIS (the World Health Organization
Library and Information Networks for Knowledge Database); the
Food and Agriculture Organization (FAO) of the United Nations’
Corporate Document Repository; and the website of the
publication New Internationalist. Finally, the LexisNexis Business
and News global database was added as a new resource for this
2008 update. This searches newspapers, wires and business
press. Full details of these searches are also provided in Annex
1.
A bibliography provided by WHO of global background material
including recent key reading on the amount and effect of food
marketing to children, government regulations, monitoring and
evaluations, nongovernmental organizations and private sector
reports and other relevant resources, was assessed for relevant
publications. Internal resources and contacts were also used to
highlight key literature.
3.3 Initial
relevance
assessment
The titles and abstracts generated through the new searches
were assessed against relevance criteria to determine whether
or not they should be included in the review (see Box 1). The
definition of promotion was extended for this update to include
advertising via digital media. Although the Internet has been
included in the criteria since the 2003 review, promotional
channels utilize other digital media such as mobile (cell) phones
and SMS or text messaging.
These criteria were applied to the studies yielded by the
database searches. This process identified an additional 52
studies on the extent and nature of food promotion to children
and an additional 20 examining its effects (see References).
These studies were added to the studies already included in the
2006 Review. This resulted in a total of 115 studies on the extent
and nature of food promotion to children and 90 on its effects.
9
Box 1: Initial relevance criteria
The following research is included:
• English or foreign language articles on the extent and nature of food promotion to children.
• English or foreign language articles on the effects of commercial food promotion on
children’s food knowledge, preferences and behaviour.
Provided that the research meets the initial relevance criteria outlined below:
• has a publication date of 1970 onwards;
• is a primary study or review;
• where the food promotion described derives from a commercial source;
• where terms mentioned correspond to agreed definitions of
- Children: those between the ages of 2–15 years
- Food: any food, non-alcoholic drink but not food supplements, vitamins or infant
formula
- Promotion: includes advertising (television, cinema, radio, print and digital media),
Internet, packaging and labelling, branding, point-of-sale material, merchandising,
film and television programme tie-in characters, and the commercial sponsorship of
education material by a commercial source
- Food knowledge, preferences and behaviour:
- food knowledge includes: perceptions of ‘good’ and ‘bad’ foods,
perceptions/understanding of a balanced diet, perception/knowledge of the
nutritional value of foods, understanding of the composition of foods, and
understanding of nutritional concepts
- food preferences includes: liking for specific foods and preferences between
different foods
- food behaviour includes: food purchasing behaviour, food purchase-related
behaviour, food consumption behaviour, and diet and health status.
3.4 Assessment
of study quality
For the studies examining the effects of food promotion, a
quantitative rating scale was developed to assess the quality of
each of the included studies that were capable of demonstrating
a potential causal relationship between food promotion and
children’s food knowledge, preferences and behaviour. Studies
were scored, using a scale of 1–5, where 1 = poor and 5 = very
good, on five criteria (see Box 2). The minimum a study could
score was five and the maximum 25. Studies were scored and
banded into three categories: 5–11 = lower scoring studies, 12–
18 = medium scoring studies and 19–25 = higher scoring
studies.
From the previous reviews, 33 studies were capable of
demonstrating a potential causal relationship between food
promotion and children's food knowledge, preferences and
behaviour, thus scored on quality in the Hastings et al. 2003
review, 5 additional effect studies were scored in the Hastings et
al. 2006 review, and in this update, 8 additional effect studies
were scored on quality.
For this update, two reviewers conducted the ratings
independently. In total, there were three discrepancies between
the reviewers, all of which were resolved through discussion and
by re-analysis of the studies with input from other academics.
These judgements of quality were then used in assessing how
much weight to attach to the findings of each study.
10
Judgements on the strength of evidence for each review
question (i.e. whether it was “weak”, “modest” or “strong”) were
based on an overall assessment of the cumulative evidence
against three criteria: the number of studies indicating a positive
effect, the size of the effect reported in the studies, and the
methodological quality of the studies themselves.
Box 2: Quality rating criteria for effects of food promotion studies
Effects of food promotion
• Only studies of effects that were capable of demonstrating a potential causal relationship
between food promotion and children’s food knowledge, preferences and behaviour were
quality assessed (n = 46). These studies were assessed on the following five criteria: the
quality of the exposure measure, the quality of the effect(s) measure(s), the
appropriateness of the analysis procedures, the extent and thoroughness of the analysis,
and the clarity and completeness of data reporting. Where a study was capable of
answering more than one specific review question (e.g. if it measured the effects of food
promotion on both knowledge and consumption behaviour), a separate rating was
obtained in relation to all relevant questions as different effects measures and analyses
may have been used.
• The remaining studies, whose results are discussed under the question ‘How do children
respond to food promotion?’, were not capable of demonstrating a causal relationship,
and were mostly simple surveys reporting only descriptive data. An initial relevance
criteria assessment was used to determine inclusion or exclusion, but quality rating
assessment of these studies was not conducted.
3.5 Other
evidence on the
extent and nature
of food
promotion in low
and middle
income countries
Separate searches were not undertaken for this 2008 update to
specifically identify evidence of the extent and nature of food
promotion in developing countries due to limited resources.
However, the updated searches detailed in Section 3.2 and
Annex 1 contained elements of the specific developing countries
searches run for the 2006 Review. New research and additional
data was identified this way and is reported in Section 4.2.
11
4. THE EVIDENCE BASE
This section presents the main review findings. Results are
grouped into five key sections. Section 4.1 describes evidence of
the extent and nature of food promotion to children from the
published literature. Section 4.2 draws on the evidence to
describe the nature and extent of food marketing in developing
countries and rapidly growing markets. Section 4.3 examines any
effect that this food promotion is having on children. The final two
sections investigate the extent of this effect relative to other
factors (Section 4.4) and the effects of food promotion on brand
and category sales of food items (Section 4.5).
4.1 The extent
and nature of
food promotion
to children
One-hundred and fifteen studies provided evidence of the extent
and nature of food promotion to children and met the inclusion
criteria for the review (see Annex 2 for a summary of the studies
included). Ninety-nine involved the collection of original data, and
the remaining 16 were review articles. Half of the included
studies (n=58) were conducted wholly in the USA. The remaining
studies were undertaken across a range of countries including
Australia, Canada, the Islamic Republic of Iran, Italy, Malaysia,
the Netherlands, New Zealand, Portugal and the United
Kingdom. Publication dates ranged from the early 1970s, many
took place in the 1990s, but two thirds were conducted during the
2000s. In the main, the studies used content analyses of
television advertising to children. This usually involved viewing
recordings (or live monitoring, prior to the invention of video) of
television channels which carried advertisements. Recordings
were often made during children’s programming, for example on
Saturday mornings or on weekdays immediately after school.
Results from the studies are briefly summarized below under the
three research questions. There is a particular emphasis on the
results of the newly included literature published between 2006
and 2008, and changes since the 2006 review (Hastings et al.,
2006):
1. What promotional channels are used by food marketers
to reach children?
2. What food items are promoted to children?
3. What creative strategies are used by food marketers to
target children?
4.1.1 WHAT
PROMOTIONAL
CHANNELS ARE
USED BY FOOD
MARKETERS TO
REACH CHILDREN?
Television continues to be the main channel used by food
marketers to reach children, certainly in terms of expenditure
(Federal Trade Commission, 2008). From the most recent wave
of literature published between 2006 and 2008 included in this
review, over half of the 52 studies examined television
advertising. However the dominance of studies of television
advertising is much diminished compared to the last review
(Hastings et al., 2006), in which only three of the studies did not
12
examine television advertising (Consumers Union, 1995;
Hawkes, 2002; Longman, 2002). Most of the television
advertising studies examined adverts that featured during
children’s programming ‘time-slots’: Saturday and Sunday
morning television and on weekdays after-school. These timeslots were shown to be heavily used by food marketers to
promote foods to children. This trend was consistent across
countries (see 4.1.2). However, a number of researchers
recognize that children still view television programmes not
specifically targeted for them (e.g. Gantz et al., 2007). Television
adverts were also recorded from broader programming time slots
and from commercial broadcast, commercial cable and noncommercial channels.
In the previous review, there was little mention of other forms of
above-the-line promotion (i.e. direct advertising) such as the
printed media (e.g. comics or magazines), on public signage,
through direct mailing or through the Internet (Hastings et al.,
2006). In the last couple of years more research has been
published in these areas. Some concentrate on one type of
media only, such as magazines (Jones & Fabrianesi, 2006;
2008; Kelly & Chapman, 2007), in-store advertising (Mazur et al.,
2008) and websites (Alvy & Calvert, 2008; Moore, 2006; 2008;
Moore & Rideout, 2007; Weber, Story & Harnack, 2006). Other
studies have included multiple media, such as magazine food
adverts and the websites the adverts direct readers to (Cowburn
& Boxer, 2007). Below-the-line promotional techniques such as
sponsorship, point-of-sale, free samples of food items, free
gifts/tokens (premiums) with food items, mobile phone
messaging, loyalty clubs, interactive food, novel packaging, tieins with movies and television programmes, tie-ins with computer
software and other forms of wider brand building were examined.
In-school marketing was identified as a useful channel for
promoting foods to children (Consumers Union, 1995;
Consumers International, 1999; Horgen, Choate & Brownell,
2001; Longman, 2002; Molnar et al., 2008). Even in comparison
to television, this type of promotion was felt to be particularly
effective at reaching children (Consumers Union, 1995).
Twelve studies from the updated 2006 to 2008 literature
investigated the Internet as a promotional channel used by food
marketers to children, and 11 of these included a content
analysis of food companies’ websites. An analysis of 44 US food
and drinks companies in 2006 found that two-thirds had website
space specifically for young people, or an independent website
for products that appeal to children, or had ‘advergames’ (an
internet-based or downloadable videogame promoting a brandname product by featuring it as part of the game) (Federal Trade
Commission, 2008). A study from the USA found nearly threequarters of food advertisers’ websites included advergames and
the same amount had brand-related content (such as wallpaper
or screensavers) the child could use after leaving the website.
Over half of the websites had television commercials to view,
and a number included polls or quizzes the company could use
for market research (Moore, 2006; 2008; Moore & Rideout,
2007). Similar techniques have been found worldwide (Lobstein
et al., 2008), in the Asia Pacific region (Robinson, 2008),
Australia (Kelly et al., 2008a), Malaysia (Ho & Len, 2008) and the
United Kingdom (Which?, 2008).
13
Only a few studies examined changes in the promotional
channels used by food marketers over time. The findings from
these studies, coupled with original advertising spend data from
AC Nielsen, confirm that television was the dominant promotional
channel throughout the previous decade. However, advertising
spend began decreasing slightly in both relative and absolute
terms from 2002. In the USA for example, in 2006, 46% of
promotions expenditure by food and drinks companies was on
television advertising, and 53% was by television, radio and print
advertising combined (Federal Trade Commission, 2008). Five
per cent of expenditure accounted for new media promotion. As
the authors state, a focus on expenditure data may
underestimate the degree to which food and beverage marketers
used the Internet to reach children and teens. Quantifying the
extent of food marketing to children on the internet the authors
calculated that display ads (e.g. banner ads) for food and
beverages generated roughly 2 billion impressions4 on childoriented websites and more than 9 billion impressions on teenoriented websites in 2006 (Federal Trade Commission, 2008).
One study looked at potential future trends in food promotion to
children including below-the-line marketing activities such as
branding, packaging, and the advent of new ‘fun’ food (Longman,
2000).
Overall, the review found that television continues to be the main
channel used by food marketers to reach children. There is
evidence that the dominance of television has recently begun to
wane, particularly as other media such as the Internet grow.
4.1.2 WHAT FOOD
ITEMS ARE
PROMOTED TO
CHILDREN?
In the 2006 review (Hastings et al., 2006), 25 studies
investigated the extent of food promotion to children by
comparing it with the promotion of other products directed
towards children. These analyses show that only toys threatened
the prominence of food in terms of advertising, and then usually
only in the run up to Christmas. Perhaps reflecting a change in
the research agenda, only one study from the 2006 to 2008 wave
of literature compared food promotion to the promotion of other
products. 2004 estimates for 2-11 year olds suggest ‘games,
toys and hobbies’ and ‘screen/audio entertainment’ each
represent around a third as much of annual exposure to TV
advertising as the share of the ‘food products’ category (Holt et
al., 2007).
Fourteen studies compared the extent of food promotion to
children with the extent of food promotion to adults and found
that food makes up a far greater proportion of promotions aimed
at children than at adults. For example, Chestnutt & Ashraf
(2002) examined nearly 250 hours of United Kingdom television
and found significantly more food advertising during children’s
programming (62.5%) than on prime-time television (18.4%).
Similarly, in Australia, Neville, Thomas & Bauman (2005)
observed that confectionary adverts were three times more likely
and fast-food adverts twice as likely to feature during children’s
programming compared with adult’s programming. However, an
Australian study found that food adverts taped from all three
commercial stations on free-to-air television occurred in similar
4
The number of times an online ad is displayed to a website visitor.
14
proportions during children’s and adults’ viewing hours (25.5%
versus 26.9% of all advertisements respectively) (Kelly et al.,
2007). The results of another survey in Australia suggested that
adults perceived distinctly different messages in magazine
adverts targeting parents compared to the adverts targeting
children and the same food products (Jones & Fabrianesi, 2006;
2008). Website analysis found viral marketing (a peer-to-peer
promotional technique) was more common on food and drink
websites for children and teens than on websites including adult
content (Moore, 2006; 2008; Moore & Rideout, 2007).
Eighty-four studies analysed children’s food advertisements to
determine which sorts of food products were being promoted.
Televised children’s food promotions were found by virtually
every relevant study to be dominated by a ‘Big Four’ of food
items: sugar-sweetened breakfast cereals, confectionary,
savoury snacks and soft-drinks. Examples include a study in
Portugal (Lemos, 2004) of 504 children’s adverts which found
26% were for breads and sugared cereals, 35% were for sweets
(chocolates and cookies) and 12% were for soft drinks. Similarly,
an American study (Kotz & Story, 1994) found that sugared
cereals were the most advertised product (23.0%), followed by
candy (15.0%), cereal with sugar as the main ingredient (10.3%),
low-sugared cereals (6.0%) and soft-drinks (5.6%). A study from
the Islamic Republic of Iran found that the most frequently
advertised food items were salty or fatty snacks (e.g. cheese
puffs, crisps), biscuits or cakes, and soft drinks (Maryam et al.,
2005). A Canadian study of ‘regular’ foods (not ‘junk’ foods) on
sale in a supermarket repackaged to appeal to children found
that 89% (of those with on-pack nutritional data) could be
classified as of poor nutritional quality (Elliott, 2008a; 2008b).
Only 16 studies looked at time trends in the types of food being
promoted to children. Three of these were published between
2006 and 2008 (Holt et al., 2007; Maher et al., 2006; Warren et
al., 2007). Adverts for fast-food outlets were found to have
‘significantly’ increased their share of children’s adverts in recent
years. The mix of categories of food adverts seen by children in
2004 was more evenly spread across all food categories than in
1977, when cereals and desserts/sweets dominated (Holt et al.,
2007). Maher et al. (2006) found no evidence that food
processors and restaurants had changed their promotional
messages during children’s TV programming slots between 2000
and 2005. Warren et al. (2007) also concluded the same
between 2004 and 2006. Advertising spend on fast-food brands
in the United Kingdom increased in both relative and absolute
terms during the previous decade, mirroring trends in the USA,
and fast foods have now displaced breakfast cereals as the most
promoted products (Hastings et al., 2003).
Overall, the review found that children’s advertising was
dominated by food products. The so-called ‘Big Five’ has
replaced the ‘Big Four’ of pre-sugared breakfast cereals, softdrinks, confectionary, savoury snacks with the steady increase in
fast-food advertising. The advertised diet contrasts strongly with
that recommended by the public health community. It is
consistently higher in fat, sugar and salt, and healthier food items
like fruits and vegetables are significantly under-represented or
absent in some markets surveyed.
15
4.1.3 WHAT CREATIVE
STRATEGIES ARE
USED BY FOOD
MARKETERS TO
TARGET CHILDREN?
Sixty-nine studies examined the creative strategies used in food
promotions to target children. A wide range of strategies were
explored including:
•
•
•
•
•
the use of characterization (e.g. gender and race) and
animation
the use of humour or other themes
the types of theme appeals or messages used to attract
children
the use of disclaimers that provide information about the
products
the use of premiums and competitions.
In terms of characterization, children’s televised food advertising
typically featured off-screen male announcers and on-screen
male characters. Other adults who appeared on-screen in food
adverts tended to be portrayed as either comic-book heroes or
villains. The use of animation techniques in television food
adverts was found to be particularly strongly associated with
children’s food adverts in comparison to non-food adverts aimed
at children and adult-oriented food adverts. For example, in one
early study, over 60% of food adverts to children used some kind
of animation, compared with only 1% of toy adverts (Atkin,
1975a/Atkin & Heald, 1977). Animation or mixed formats were
also seen to highlight the light-hearted or humorous tone of
children’s food adverts; much less humour was observed in both
non-food adverts aimed at children and adult-oriented adverts.
The Batada et al. (2008) analysis of Saturday morning television
food advertising found the most commonly used techniques
continued to be movie, cartoon, animated or costumed
characters (used in 74% of food adverts). Conversely, another
television study conducted in the USA (Gantz et al., 2007)
evaluated all programming during a set time period and found
that a relatively small 11% of food adverts targeting children or
teens used a children’s TV or movie character.
Theme appeals in children’s adverts were examined in 18
studies in the Hastings et al. (2006) review. A further ten studies
published between 2006 and 2008 were included in the updated
review. Although there is little consensus about the definition of
these themes, the review shows that food marketers are
generally using appeals based on the following:
•
•
•
•
•
•
•
taste (e.g. emphasizing sweetness)
nutritional or health properties
the physical appearance or texture of the food
fantasy and adventure themes
fun and humour
price
novelty
The most popular appeals used in the promotion of foods to
children are hedonistic, including taste, humour, actionadventure and fun. For example, an early American study found
that although 98% of toy adverts used a serious appeal, over
90% of food adverts used a humorous appeal (Atkin, 1975a/Atkin
16
& Heald, 1977). Similarly, a more recent study of advertising in
the Netherlands found that nearly 90% of children’s food adverts
used taste as an appeal and around 85% used humour (Buijzen
& Valkenburg, 2002). In the Islamic Republic of Iran, one study
found that 56% of food adverts to children used ‘taste’ appeals
(Maryam et al., 2005). In general, breakfast cereal adverts alone
were found to regularly use nutritional appeals, regardless of
whether or not these appeals were deemed to be misleading
(Lobstein et al., 2008; Ho & Len, 2008).
Twelve studies examined the nature of disclaimers. Other
products advertised to children, such as toys, were much more
likely to use disclaimers than were food items and services,
although the chief exception to this pattern was breakfast
cereals. Where disclaimers were used, adverts for products like
breakfast cereals, candy and gum tended to use informative
disclaimers (referring to what the product does do), while fastfood restaurants tended to use restrictive disclaimers (referring to
what the product does not do) (Muehling & Kolbe, 1998). The
recent study by Gantz et al. (2007) into television food
advertising to children in the USA found that 22% of food
advertising to children or teens included a disclaimer (e.g. “part
of a balanced diet” or “enjoy in moderation”).
Some research suggests that adverts designed to promote food
to children utilize ‘pester-power’ or ‘purchase-influence-attempts’.
A commonplace creative strategy employing pester-power
identified in the literature is the use of premiums or competition.
Prizes include collectibles (e.g. toys), and the use of celebrity
figures (although only limited cases). Moore’s study of food
marketing to children online found that 38% of websites used
incentives for product purchases such as access to videogaming tips after keying in a code from a bought food package
(Moore, 2006; 2008; Moore & Rideout, 2007). Six studies
examining programme–commercial tie-ins found the boundary
between television shows and advert breaks to be unclear. The
food products mainly promoted through show sponsorship or
utilizing tie-ins, tended to be those categorized as pre-sugared or
of low nutritional value (e.g. Consumers International, 1996;
Hawkes, 2002).
Time trends, examined in ten studies, suggest that the basic
creative strategies used to promote food to children are
beginning to change. The increasing use of new media is giving
rise to a host of new potential creative strategies. Social
networking, viral marketing, instant messaging, virtual worlds
(Chester & Montgomery, 2007; 2008), and the evolution of
brand-stretching and ‘globalization’ is allowing promotional
messages to cut across many different media and also allowing
increased tie-ins with below-the-line marketing activities (Which?,
2008; Lobstein et al., 2008).
This review found that themes of fun and fantasy, or taste and
flavourings, rather than health and nutrition, continue (as in the
2006 review) to be used to promote food to children.
17
4.1.4 CONCLUSION
From the research evidence, children’s food promotion has been
dominated by television advertising over the past few decades.
The majority of this promotes the ‘Big Five’ group of food
products, namely pre-sugared breakfast cereals, soft-drinks,
confectionary, savoury snacks and fast-food outlets. There is
some evidence that the dominance of television has begun to
diminish particularly with the rise of digital media. The
importance of strong, global branding reinforces a need for
multifaceted communications combining television with
merchandising, ‘tie-ins’ and point-of-sale activity.
The advertised diet contrasts sharply with that recommended by
public health advisers. Themes of fun and fantasy, or taste, are
used to promote to children. Health and nutrition is not a
common theme for promotion, and the recommended diet for
long term good health gets little marketing support.
4.2 Food
promotion and
marketing in
developing and
middle income
countries
This section examines the main strategies and channels used by
food marketers in developing countries and ‘emerging
economies’ – countries where increased openness of markets
and/or increased personal disposable incomes has been
followed by increased activity in food marketing, particularly of
added-value foods and globally traded products. The Hastings et
al. 2006 review conducted additional desk research on food
marketing in developing countries and rapidly growing markets
using a variety of data sources (e.g. business and marketing
press and journals and commentaries from non-government
organizations) to highlight the interaction of modern food
marketing strategies and techniques with local cultures and
socioeconomic change. Children and young people play a critical
role in this change process, and represent an important target
market. Research from low and middle income economies found
using the search strategies detailed in Annex 1 updates this.
4.2.1 WHAT
MARKETING
STRATEGIES ARE
BEING USED TO
PROMOTE FOODS IN
DEVELOPING AND
MIDDLE INCOME
COUNTRIES?
Four marketing strategies were identified:
(i) Widespread adoption
of international
marketing and
promotion techniques
Promotional techniques and channels used to market food to
children and young people in developing and middle income
countries include television advertising, on-pack promotions,
cartoon characters and animation, interactive websites,
sponsorship of school activities, sports and competitions.
Robinson (2008) conducted a cross-sectional content analysis of
food product marketing in the Asia Pacific region. The extensive
use of sponsorship, advertising, price incentives, free gifts and
(i) widespread adoption of international marketing and
promotion techniques
(ii) marketing strategies that target children and families
(iii) adaptation of products and services to local markets
(iv) promotion of a sub-optimal dietary profile
18
on-pack promotions, celebrity endorsements, movie tie ins,
language teaching, clubs and party offers, school targeting by
the six multinational food and drink companies with greatest
market share in the region was reported.
Building brand loyalty
Some techniques, such as free club memberships, and strong
logos and characters are particularly effective in building brand
value and loyalty. Hawkes (2002) noted Pizza Hut statements on
instilling brand loyalty in young children through the use of
animated advertisements (Advertising Age, 2000, as cited by
Hawkes, 2002). In China, Kentucky Fried Chicken (KFC) created
Chicky, a cartoon, baseball-cap-wearing chicken to replace the
traditional Colonel Sanders figurehead (Watson, 2000a). Ho &
Len (2008) audited the marketing of breakfast cereal products in
Malaysia and found extensive use of a brand building
techniques, such as sponsorships, cartoon characters and
animation.
Television advertising
Television remains a key channel for marketing food to children
in many countries. Vignali (2001) stated that the success
experienced by McDonald’s in East Asia could not have been
achieved without appealing to children and teenagers though
television advertising. In 2004, Pizza Hut spent approximately
$US 2.5 million on television advertising in India (Exchange for
Media, 2004). In Thailand, TV advertising was identified as the
most important medium for snack food advertising to the young
(Mulchand, 2004, as cited by Hawkes, 2006). Atktas Arnas
(2006) analysed advertising content of children’s TV
programming in Turkey and found 344 out of 755 adverts were
for food products. Galcheva, Iotova & Stratev (2008) found 33%
of advertising during children’s TV time in Bulgaria was for food
and beverages. Karupiah (2008) found advertising of food
products in Malaysia was most intensive during children’s prime
time viewing periods, including holidays, school vacation time
and weekends. Temple, Steyn & Nadomane (2008) found 17%
of South Africa TV advertising during children’s programming
was for food and beverages.
Sports stars and celebrities
Sponsorship strategies combine the support of global events
such as the Olympic Games and the FIFA (Fédération
Internationale de Football Association) World Cup, where
internationally-known pop celebrities and sports stars cut across
national boundaries, with local activity. For example, Hawkes
(2002) notes KFC sponsorship of basketball, a sport with many
young fans in developing countries, and badminton in Malaysia
where it is particularly popular. It seems to be effective: in their
study of brand recall from cricket advertising, LODESTAR
attributed PepsiCo’s higher recognition status in India, than
Coca-Cola, to its long established association with the nation’s
passion, cricket. Endorsement by famous cricketers and its high
visibility at matches appears to be more effective than CocaCola’s greater TV advertising spend (LODESTAR, 2002).
19
Collectable toys
Child-oriented distribution,
including targeting of
schools
Fast-food retailers have child-specific menu choices, such as the
McDonald’s ‘Happy Meal’ and KFCs ‘Chicky Meal’, and a
mainstay of these is a ‘free’ toy which is typically part of a
collectable set and therefore encourages several restaurant
visits. This strategy has been particularly successful in countries
where historically there has been a scarcity of manufactured
toys. Hawkes notes their presence has generated massive
buying activity and even sit-ins when the supply of toys ran out
(Hawkes, 2002).
Hawkes (2002) charts the expansion of soft drinks manufacturers
into schools in South America with the use of promotional gift
packs. This targeting of schools is credited with raising
consumption by schoolchildren by 50% in Costa Rica (Panamco,
1998, as cited by Hawkes, 2002). Focusing on the connection
with ‘place’ she notes how buying opportunities have been made
available in or near schools or other areas where teenagers
congregate, such as cinemas and cafés. Ho & Len (2008)
describe the use of school programmes in Malaysia to promote
breakfast cereals, Robinson (2008) also found sponsorship of
school sports in the Asia Pacific region an integral part of the
marketing strategies of multinational food and drink companies.
Interactive websites
A survey of 32 low, lower middle, and upper middle income
countries (Lobstein et al., 2008) and the Asia Pacific survey
(Robinson, 2008) both note the use of interactive websites
offering entertainment such as games, puzzles, cartoons, as well
as free offers such as ring tones and competition prizes. The
authors report that sites did not require parental consent.
(ii) Marketing strategies
that target children and
families
Creative strategies known to attract and engage children in the
developed world, and described in Section 4.1 are found to be
similarly employed in lower income countries. There is strong
emphasis on fun, fantasy, excitement as well as novelty in taste
and appearance of foods and ingredients (Robinson, 2008;
Lobstein et al., 2008; Galcheva, Iotova & Stratev, 2008; Ho &
Len, 2008).
Additionally, Ho & Len (2008) and Lobstein et al. (2008) describe
marketing appeals to parents of child-oriented products. These
messages emphasize nutrition and growth, academic and
sporting success and family harmony. Parents may therefore be
simultaneously experiencing both health and nutrition
reassurance messages from manufacturers and retailers, and
demand stimulated by creative messages for the marketed
products from their children.
Marketers are recognizing and responding to the growing
importance of children as purchase influencers in developing
countries. This is particularly marked in China, which has seen a
shift from children as the passive recipients of food chosen by
their parents/caregivers in the 1960s to a 1990s environment
5
where a Beijing study credited children with determining 70% of
household spending. McNeal & Yeh (1997) found that in the 4–
10-year-old age group, food stores were a frequent target for
5
Research by James McNeal and colleagues in Beijing, cited in Asiaweek, 1 Dec., 1995
20
children’s considerable influence on their parents’ purchasing.
Ethnological studies by Watson record the important part played
by children in the acceptance of western style fast-food chains in
the late 1980s. For example, fast-food outlets empowered
Chinese youngsters to act as cultural guides to the fast-food
experience for their parents and elders (Watson, 2000a). China’s
adoption of a one child policy in 1979 has resulted in a ‘4-2-1
indulgence factor’ where four grandparents and two parents
focus their attention and money for treats on one child. This
money and children’s own spending money is frequently spent in
fast-food outlets (McNeal & Yeh, 1997). The market potential of
children in China is enormous. Average child-related
consumption was estimated in 2003 to represent one third of
household consumption in 85% of households (Ying, 2003, as
cited by Chan, 2005).
Other studies note how marketers have identified children in
developing and emerging economies as consumers in their own
right. McNeal & Yeh (1997) describe how McDonald’s and Pizza
Hut have relied heavily on children to expand their overseas
market in both Europe and Pacific Asia.
Watson (2000a) notes that a ‘fun’ atmosphere, child-oriented
brightly-coloured picture menus and the partial withdrawal of
adult supervision helped counteract the traditionally subordinate
position of the young in Chinese culture, and promote their
freedom to choose. Similarly, Vignali (2001) notes that
McDonald’s mascot Ronald was paired with a female Aunt
McDonald in Beijing, and that each restaurant assigns between
five and ten female assistants to take care of children and talk to
parents.
As noted above, the marketing of dairy products is substantial
(Aktas Arnas, 2006; Karupiah 2008; Temple, Steyn &
Nadomane, 2008). In many of these countries milk is not a
traditional foodstuff, and the marketing activity noted in these
studies may reflect corporate efforts to build new product sectors
through children and their families.
(iii) Adapting products
and services to local
markets
Food marketers aim to create a standardized product, but
recognize that the ability to adapt to local environments is also
very important: they ‘think global, act local’ (Vignali, 2001). This
affects both the products they develop and the service
environment in which these are delivered.
Taste, for instance, is a crucial characteristic of food products,
and a study by the American Agricultural Trade Office aimed at
promoting American trade noted that suppliers should be aware
of the specific taste factors that are successful in winning
customers (Baker, 2000). Similarly Vignali notes incidences of
McDonald’s having to adapt to the needs of specific religious
laws, such as abstinence from eating beef by Hindus and pork by
Muslims. This adaptation goes beyond religious observance to
cater for local tastes, with Teriyaki Burgers sold in Japan and
Samurai Pork Burgers with sweet sauce meeting the taste
requirements of customers in Thailand (Vignali, 2001). Pizza Hut,
the world’s largest pizza chain with over 12,500 restaurants
across 91 countries first entered India in 1996. More vegetarian
21
pizza recipes were developed and a number of vegetarian-only
outlets opened, the only ones in the world.
Vignali (2001) also notes that the developed country style of fast
food is viewed as a high quality product. Lee & Ulgado (1997)
utilized a perceived service value questionnaire (measurement of
the gap between expectations and perceptions of performance
levels) to test for differences in American consumers and those
in the Republic of Korea. Low prices were seen as paramount to
the American consumer whereas customers in the Republic of
Korea placed more weight on service dimensions such as
reliability and empathy. The authors also note that while
Americans put the emphasis on fast food, Asians see eating,
especially at a restaurant, as a more social, family-oriented
experience (Copelands, 1985, as cited by Lee & Ulgado, 1997;
Hall, 1966, as cited by Lee & Ulgado, 1997).
Witkowski and colleagues (2003) of the California State
University compared the brand identity of KFC in China and
America by surveying 795 students and concluded that Chinese
consumers valued different attributes. For example, they found
that fast-food restaurants provided a safe environment for Asian
women, because they are both alcohol and smoke free (Watson,
1997, as cited by Witkowski, Ma & Zheng, 2003). (KFC is one of
the few restaurants in China that does not permit smoking
according to Iritani, 2001, as cited by Witkowski, Ma & Zheng,
2003.) Compared to their American counterparts, Chinese
consumers were more likely to eat within fast-food restaurants
and to stay longer. Reasons suggested include a lack of cars,
which militate against drive-through facilities; a culture of slower
eating; and the higher relative cost of eating at a fast-food
restaurant encouraging an inclination to linger. Furthermore,
global company outlets are perceived to offer better standards of
hygiene and a more relaxed atmosphere than indigenous
retailers. Overall, Chinese respondents reported a more positive
brand impression of KFC than their American counterparts
(Witkowski, Ma & Zheng, 2003).
Food marketers also attract market share with product and
service positioning that offers modern, international eating
experiences. Being seen to eat in global fast-food chains can
enhance status. Anderson & He (1999) cite Lu’s 1994
observation, that Chinese teenagers did not like pizza, but would
visit Pizza Hut to be seen. Watson (2000b) noted that during
McDonald’s early years in China, it was promoted as an outpost
of authentic American culture, offering authentic American
hamburgers. Fast-food outlets can become the place to be seen,
offering the food to be seen eating for an age group noted for
their susceptibility to peer pressure (Wiener, 2004).
Watson (2000a) has also observed that marketing innovations
can trigger cultural change. For example, the introduction of
events and services marking traditional western celebrations.
Solar birthdays, Thanksgiving, Halloween, Fathers’ and Mothers’
Day are examples of these previously unknown special
occasions that fast-food restaurants have actively introduced to
developing economy cultures and have in turn generated
important marketing and positioning opportunities. Watson
(2000a) notes that birthday parties became so popular in Beijing
that KFC had to rapidly build special reserved areas in the
22
restaurant. McDonald’s birthday parties for children were led by
Aunt McDonald introducing party-goers to singing ‘Happy
Birthday’ and other novel experiences.
(iv) Promotion of a suboptimal dietary profile
The cross-sectional surveys of food and marketing activities in
32 low or middle income countries by Lobstein et al. (2008),
and the Asia Pacific region by Robinson (2008), found food
promotion was heavily concentrated on high fat, salt and/or
sugar, with particular focus on salty snacks, sweets/candy and
chocolates, sweetened breakfast cereals and soft drinks.
Similarly, content analysis of TV advertising found the emphasis
was very heavily concentrated on energy dense, low nutrient
foods. In Bulgaria, Galcheva, Iotova & Stratev (2008) found the
most commonly advertised products were salty and sweetened
snacks and cereals, sweets, soft drinks and juices and salty
foods and no commercials for fruit or vegetables. In Malaysia,
Karupiah (2008) reported the most frequently advertised foods
on children’s TV were snacks (34.5%), dairy products (20.3%),
sugars and candies (13.4%), biscuits (11.2%), fast food (6.7%),
breakfast cereals (6.4%) , beverages (4.1%) as well as a range
of other miscellaneous processed products. No fruit or
vegetable advertisements were reported. In South Africa,
Temple, Steyn & Nadomane (2008) found 55% of TV food
advertising was for foods of ‘poor nutritional value’ and there
were no advertisements for fruit or vegetables.
The advertised diet in the low and middle income countries
surveyed therefore contrasts sharply with recommended
optimal dietary patterns which emphasize generous quantities
of fruit, vegetables, complex, unsweetened carbohydrates and
moderate intakes of lean protein and calcium sources.
There is evidence that children do not understand the nature of
advertising. Chan (2000) tested 448 children aged 5–12 years
in China, Hong Kong Special Administrative Region (SAR), and
found that only half of those aged 11–12 years understood that
television stations get paid to broadcast commercials.
4.2.2 CONCLUSION
Although data are sparse it is clear children in developing and
emerging economies are being widely and effectively targeted
by energy-dense food promotion, using the same techniques as
in developed economies. Cultural and political developments
have enabled and facilitated this initially, but marketing and
promotion have developed and accelerated the change
process. Children are a key target group, both as consumers in
their own right, and as a bridgehead into wider society.
23
4.3 The effects of
food promotion
to children
4.3.1 HOW DO
CHILDREN RESPOND
TO FOOD
PROMOTION?
This section examines evidence of the effects of food promotion
on children. Effects on behaviours (purchase, or purchase
requests, consumption patterns), potential determinants of
behaviour (preference, attitudes, knowledge, beliefs) and dietrelated health outcomes (adiposity, physiological indicators etc)
were all considered. Section 4.3.1 reviews the evidence provided
by observational and correlation studies, and 4.3.2 reviews
studies designed to examine causality. A summary of all the
studies included in this section is provided in the data extraction
tables in Annex 3.
This section provides an overview of narrative studies and
studies that explored the qualitative nature of the associations
between food marketing and children’s responses. Forty studies
are included in this section of the review. Sixteen of these
studies were conducted in developing countries including Chile,
India, the Islamic Republic of Iran, Saudi Arabia, the Bolivarian
Republic of Venezuela and Puerto Rico. A large number of
studies were conducted in the USA. Research identified from
New Zealand, Australia, Turkey, Canada and the United
Kingdom is also included here.
The studies examine nine different types of response to food
promotion: recall of food advertising, liking for and attitudes
towards food advertising, communication about food advertising,
purchase-request behaviour perceived to be triggered by food
promotion, responses to free gifts and packaging, food
preferences, food purchase behaviour, and food consumption
behaviour. Differential responses according to age, gender,
socioeconomic status and race were reported in some studies
and in many of the studies these factors were recognized as
potentially confounding variables. However, no consistent pattern
in the analysis of potential confounding factors was apparent and
therefore it was not possible to draw any conclusions on the
pattern of interactions, except to note that these were recognized
in the research but require more robust control and analysis in
future research.
Many studies found strong recall of food adverts by children.
Hitchings & Moynihan (1998) found that 9–10-year-old English
children were able to recall seeing adverts in the past two weeks
in seven different food product categories. Barry & Hansen
(1973) found second grade black and white children could recall
advertising content. Batada et al. (2008) found half of a sample
of 58 children could accurately match from memory, and without
prompting at least 5 of 10 logos/characters with TV
advertisements for breakfast cereals. Studies from the
developing world produced similar results and also reported that
food adverts tended to be among children’s favourites. For
example, Olivares et al. (1999) found that 80% of children
interviewed in Chile were able to recall food adverts that they
liked.
Carruth, Goldberg & Skinner (1991) and Yavas & Abdul-Gader
(1993) found that children discussed food promotion with peers
and families. Lam (1978); Del Toro & Greenberg (1989); Yavas &
Abdul-Gader (1993) found children asked their parents to buy
24
food they had seen advertised. A further nine studies found that
parents believed their children were influenced by food
promotion to request specific foods and most acceded to these
requests at least sometimes. For example, Musaiger et al. (1986)
found children requested food products that they had seen
advertised and mothers in lower socioeconomic groups were
more responsive to children’s requests than more affluent
mothers. Aktas Arnas (2006) surveyed 348 mothers and found
33% reported their children requested food products advertised
on TV while viewing and 40% requested products during
shopping trips, with 9% reporting that refusal could provoke
arguments or crying. Chamberlain, Wang & Robinson (2006)
examined children’s screen media time and food requests in a
prospective study which treated screen time as a proxy for TV
food advertising exposure. The association between baseline
screen viewing and requests for advertised foods at baseline and
7-20 months later was statistically significant, after adjusting for
socio-demographic factors.
Four studies observed that free gifts and packaging attributes
appeared to attract children’s attention and stimulate demand for
products (Atkin, 1975b/1978; Donohue, 1975; Carruth et al.,
2000; Folta, Bourbeau & Goldberg 2008). Olivares et al. (1999)
and Olivares, Yáñez & Díaz (2003) reported that TV advertising
of promotions and free gifts prompted purchase and nearly 65%
of children interviewed said they continued to purchase these
products even when the promotional offer had ended. The
authors comment that children’s purchase choices are consistent
with their advertising preferences and for the food products most
frequently advertised on television. Most popular purchases were
snacks, fizzy drinks/juice and yoghurt or milk products. Folta,
Bourbeau & Goldberg (2008) also found that free gifts increased
liking for food products, despite the study sample (average age
over 9
years) expressing scepticism of marketing
communications.
Several studies explored marketing and children's preferences
for food products. Olivares et al. (1999) asked children in Chile
about their favourite types of food ‘in accordance with their
favourite food adverts’. Favourite types of food were snacks
(56%), ready meals (22%), home-made meals (22%) and fizzy
drinks (just less than 75%). Dixon et al. (2007), in a combination
of experimental and cross-sectional research, and Marshall,
O'Donohoe & Kline (2007), who used a combination of
secondary analysis and qualitative methods, all found TV
advertisements increased liking and acceptability of advertised
products. Moya de Sifontes & Dehollain (1986) interviewed
mothers in the Bolivarian Republic of Venezuela, who reported
that children’s preferred foods were chocolate drinks, cereals,
jelly, pork, sausages and ice cream. The study also found a
significant association between the amounts of time children
spend watching advertising, and their preferences for
commercial foods advertised in the media.
Carruth, Goldberg & Skinner (1991) found 8% of North American
students reported that seeing a food advert made them want to
get something to eat ‘every day’. Sixty six per
cent of the sample reported similar responses less frequently
and 27% ‘never’. Similarly, qualitative research and secondary
analysis of data by Marshall, O'Donohoe & Kline (2007) found
25
children reported watching food adverts made them ‘feel hungry’
and increased purchase desire/intent.
Purchase behaviour was examined by Olivares et al. (1999) who
found that nearly three-quarters of children said that they
purchased food or drink products advertised on television with
offers of prizes or free gifts. This effect was greater among
children from low and middle socioeconomic groups. Olivares,
Yáñez & Díaz (2003) found that 34% of children ‘always’ had the
money to buy whatever food and drink products they wished and
64% said they ‘sometimes’ had the money. Most popular
purchase choices were sweet or salty products, and drinks. An
Iranian study reported that over 90% of interviewed students
reported they selected foods “under the influence of advertised
products” (Maryam et al., 2005). Olivares et al. (1999) asked
children if they had consumed food or drinks that had been
advertised the previous day. Half of 6–8 year olds and two-thirds
of 9–11 year olds said that they had consumed advertised
products. In a later study by Olivares, Yáñez & Díaz (2003), 40%
of children interviewed reported that they had consumed sweet
or salty products advertised.
Carruth, Goldberg & Skinner (1991) and Aktas Arnas (2006) both
found foods consumed while watching TV were largely those
most commonly advertised on TV. Secondary analysis of crosssectional data by Utter, Scragg & Schaaf (2006), and a large
number of similar studies, have found TV viewing time was
positively associated with overall consumption of foods most
commonly advertised on TV.
However, without some test of validity of TV viewing as a proxy
for advertising exposure, these observational studies provide
useful contextual data and highlight areas for future research, but
do not demonstrate exact correlation of TV viewing and exposure
to TV advertising.
Any impact of food consumption patterns is likely to be
moderated by parents and other caregivers. Marshall,
O'Donohoe & Kline (2007) and Aktas Arnas (2006) for example
both found parental control of children’s purchase and
consumption preferences to be important. However, as outlined
above, many children have the financial means to make
purchase and consumption decisions. For example, in Olivares,
Yáñez & Díaz (2003), 34% of children ‘always’ had the money to
buy whatever food and drink products they wished and 64% said
they ‘sometimes’ had the money independently of their
parents/caregivers.
The studies examined in this section indicate that the creative
strategies examined in Sections 4.1 and 4.2 have persuasive
power. In all the studies that include a note on the content of the
food promotion explored in the research, content was
predominantly or exclusively for energy dense, processed food
products. Children were found to be aware, and to enjoy food
promotions. A number of studies found an association between
exposure to food marketing and/or food promotion channels with
food preferences, purchasing and consumption patterns of the
promoted food products.
26
4.3.2 IS THERE A
CAUSAL LINK
BETWEEN FOOD
PROMOTION AND
CHILDREN'S FOODRELATED
KNOWLEDGE,
ATTITUDES,
BEHAVIOURS AND
HEALTH STATUS?
To answer this question, and in view of the complexity of both
the environment in which food promotion occurs and possible
food-related intrapersonal responses which may moderate effect,
Bradford-Hill’s (Bradford-Hill, 1965) principles on inferring
causality were used to guide the review process. The approach
therefore aimed to establish if the cumulative evidence
demonstrated consistency in direction and scale of any
correlation found between food promotion and food and health
related outcomes across multiple methodologies. In addition, all
the research included here met at least one of the following
criteria: that exposure preceded effect; that the pattern of the
association was
incremental/dose-responsive;
that
the
association demonstrated specificity between the independent
variable (i.e. food promotion) and the dependent variable (i.e.
food-related individual outcome measure); that direction of
effects measures were consistent across different sample groups
or methods; that on withdrawal of the exposure, the exposureeffect association was reversible.
Experimental studies can provide direct measures of the
outcome effects, in response to food promotion while controlling
for potential confounding factors and bias, but findings and
conclusions
inevitably
have
limited
generalizability.
Observational and cross-sectional studies offer stronger
ecological validity but their real-world setting means control for
selection bias and confounding variables is less strong. The
strength or significance of the findings from whichever research
methodologies are used is a function of the computed statistical
significance, the research design, contextual influences, as well
as the clarity and accuracy of the reporting of findings and
conclusions. Therefore all the papers reviewed here were first
screened for their putative ability to infer causality from
measured associations using the Bradford-Hill guidelines, and in
a second stage, the strength of evidence was reviewed against
five criteria of the robustness of the research design, analysis
and reporting (see Section 3.4, Box 2), as well as the strength of
the statistical test results.
Forty six studies were included on the basis that their design
could test causality. All of these were conducted in developed
economies, mainly North America and Europe. Twenty nine were
experimental studies, one was quasi-experimental, thirteen were
cross-sectional and three were observational.
Each of these studies was then judged against the five quality
criteria described in Section 3.4 and the results of the statistical
analysis.
The results are discussed under the following question headings:
(i) Does food promotion influence children’s nutritional
knowledge?
(ii) Does food promotion influence children’s food
preferences?
(iii) Does food promotion influence children’s food
purchasing and purchase-related behaviour?
(iv) Does food promotion influence children’s food
consumption behaviour?
(v) Does food promotion influence children’s diet and
health-related variables?
27
(i) Does food promotion
influence children's
nutritional knowledge?
Nine studies investigated the influence of food promotion on
children’s nutrition knowledge. Five were randomized control
design, and four were cross-sectional surveys. Two studies were
higher scoring and seven were medium scoring. With one
exception all were conducted with North American respondent
samples. Most were conducted during the 1970s and 1980s.
Five studies found food promotion had an effect on, or was
associated with, differences in nutrition knowledge. Two were
experimental (Ross et al., 1980/1981; Peterson et al., 1984) and
three were cross-sectional (Wiman & Newman, 1989; Gracey et
al., 1996; Harrison & Marske, 2005). Three of the five studies
provided evidence that exposure to food promotion for ‘low
nutrition foods’ was associated with poorer nutrition knowledge.
One study found that television viewing had a significant
detrimental effect on nutrition knowledge in relation to foods that
are heavily marketed as ‘diet’ foods (Harrison & Marske, 2005).
Peterson et al. (1984) found that exposure to adverts for foods
“high in nutritional value” increased nutrition knowledge, although
it was not possible to separate out the effects of the adverts from
other nutrition messages in the study.
Three studies found no association between exposure to food
promotion and children’s perceptions of the healthfulness of
different foods or what constitutes a healthy diet. Two of these
were experimental (Goldberg, Gorn & Gibson, 1978a/1978b
Study 1; Goldberg, Gorn & Gibson, 1978a/1978b Study 2) and
one was cross-sectional (Atkin, 1975c). The study by Galst
(1980) produced inconclusive results.
In two of the studies (one which found an effect and one where
the results were inconclusive) it was not possible to separate out
the effects of advertising from other exposure variables (Galst,
1980; Peterson et al., 1984). Furthermore, studies which found
effects tended to use more detailed knowledge measures than
did the studies which did not find effects. It is possible, therefore
that the studies were not measuring the same effect.
The weight of evidence suggests that food promotion may have
little influence on children’s general perceptions of what
constitutes a healthy diet, but that it can, in certain
circumstances, have an effect on specific types of nutrition
knowledge. Overall, the evidence therefore was assessed as
modest rather than strong.
(ii) Does food promotion
influence children's food
preferences?
Eighteen studies investigated the influence of food promotion on
children’s food preferences. Seventeen were experimental, and
one was a cross-sectional study. The studies covered a wide age
range, including the 2–15 years age range examined for this
review, and some also included young people up to 18 years).
The majority of the studies were conducted in North America
during the 1980s.
Ten of the studies found that exposure to food promotion had an
impact on, or was associated with significant changes in
children’s food preferences (Goldberg, Gorn & Gibson,
1978a/1978b Study 1; Gorn & Goldberg, 1980a; Heslop & Ryan,
28
1980; Stoneman & Brody, 1981; Kaufman & Sandman, 1983;
Norton, Falciglia & Ricketts, 2000; Borzekowski & Robinson,
2001; Robinson et al., 2007; Chernin, 2008; Halford et al.,
2008b). Three of these were rated as high quality experimental
studies (Goldberg, Gorn & Gibson, 1978a & 1978b; Stoneman &
Brody, 1981; Kaufman & Sandman, 1983). The earlier studies
found children were significantly more likely to prefer high fat,
salt or sugar foods over lower fat, salt or sugar alternatives after
exposure to food adverts. The Robinson et al. (2007) study found
a direct preference for branded products over identical but
unbranded products in a randomized, blinded case-controlled
trial. Three studies found children were more likely to choose an
advertised brand than a non-advertised brand of the same
product type after exposure to food adverts (Gorn & Goldberg,
1980a; Heslop & Ryan, 1980; Borzekowski & Robinson, 2001).
One cross-sectional study found a weak association between
television advertising and preferences for specific foods (Norton,
Falciglia & Ricketts, 2000).
Six studies (five experimental and one cross-sectional study) did
not find statistically significant associations (Goldberg, Gorn &
Gibson, 1978a/1978b Study 2; Ritchey & Olson, 1983; Clarke,
1984; Peterson et al., 1984; Gorn & Florsheim, 1985; Neeley &
Schumann, 2004).
Overall, the stronger studies found evidence of significant effects
and the less robust studies did not. Cumulatively, therefore the
research reviewed indicates modest strength evidence that food
promotion influences food preferences.
(iii) Does food
promotion influence
children's food
purchasing and
purchase-related
behaviour?
Eight studies which examined the impact of food promotion on
children’s food purchasing and purchase-related behaviour were
reviewed. Purchase-related behaviour was defined as behaviour
intended to influence parents’ food purchasing selections. Three
were randomized controlled experimental studies, one was a
natural quasi-experiment, two were observational studies, and
two were cross-sectional surveys. In terms of quality, four were
higher scoring, three were medium scoring, and one was lower
scoring.
Seven studies found exposure to food promotion was
significantly associated with specific purchases or purchaserelated behaviours measured in each study. Bridges & Briesch
(2006) observational study did not find an association between
point of sale promotions or a proxy measure of exposure to
advertising and household purchase of child-oriented products.
One experimental study (French et al., 2001) found that
promotional signage on vending machines significantly increased
sales of low fat snacks in secondary schools. This effect
persisted, with and without price reductions. This study provided
the most robust evidence for a causal link between promotion
and actual purchasing behaviour by children. One study was a
natural experiment (Goldberg, 1990). This compared the
household purchase of breakfast cereals between Englishspeaking children in Quebec, exposed to American television,
and French-speaking children who had access to American
television but tended to watch more French-language television.
At the time of research (1980), French-language Quebec TV
29
banned advertising targeting children; and therefore French
speaking children were less likely to be exposed to advertising
for children’s cereals. Regression analysis found exposure to
American television was significantly associated with increased
household purchase of advertised cereals, independently of
income and language variables.
Two experimental studies found that exposure to food promotion
increased children’s purchase behaviour observed in a natural
setting (supermarket shopping with parents) (Galst & White,
1976; Stoneman & Brody, 1982). The former study also found
that the more attentive a child was to television advertising,
relative to television programmes, the greater the number of
attempts to influence parental shopping purchases at the
supermarket. One observational study (Reeves & Atkin, 1979)
and one cross-sectional study (Atkin, 1975c) also found
significant associations between amount of Saturday morning
television viewed and frequency of food purchase requests to
parents, with ‘heavy’ viewers in both studies making more
requests than ‘light’ viewers. The Taras et al. (1989) crosssectional study found a weak association between television
watching in general and food purchase requests to mothers.
Overall, the studies provide strong evidence that food promotion
influences children’s food purchase-related behaviour. Both the
methodologically stronger and weaker studies found evidence of
effects. One study, found a positive association between
purchase of low fat snack sales and low fat food promotion.
Other studies identified effects in the direction of increasing
purchase requests for foods high in fat, sugar or salt.
(iv) Does food
promotion influence
children's food
consumption
behaviour?
Eighteen studies investigated the effects of exposure to food
promotion on children’s food consumption behaviour.
Consumption behaviour was defined as including consumption of
food on a single occasion, daily selection of foods for
consumption over a short period of time, and self-reported
patterns of consumption behaviour. Twelve studies used
experimental designs, five were cross-sectional studies and one
was a prospective observational study. Two studies scored high
on the quality rating, 13 were medium scoring, and three were
lower scoring. Fourteen studies were North American and four
were European studies.
Six experimental studies using TV advert exposure as the
treatment variable found food promotion had a significant effect
on children’s consumption behaviour. The effects included
increased frequency of selecting less healthful food in preference
to healthier options (Gorn & Goldberg, 1980b/1982); increased
consumption of calories (Jeffrey, McLellarn & Fox, 1982 Study 2/
Fox, 1981; Halford et al., 2007; 2008a; Wiecha et al., 2006), and
total increased food intake (Halford et al., 2004; 2007; 2008a).
Five cross-sectional studies (Atkin, 1975c; Bolton, 1983; Ritchey
& Olson, 1983; Boynton-Jarrett et al., 2003; Buijzen, Schuurman
& Bomhof, 2008) found small associations, of varying degrees of
strength, between exposure to television food advertising (as
measured using television viewing) and frequency of snacking or
consumption of specific foods. The studies were of varying
quality. Boynton-Jarrett et al. (2003) found that for each
30
additional hour of television children viewed per day, daily
consumption of fruit and vegetables was reduced by 0.14
servings. Buijzen, Schuurman & Bomhof (2008) found the
association between consumption of branded products
advertised on TV was moderated by income and by parental
communication styles.
A prospective observational study (Wiecha et al., 2006) found a
positive association between TV viewing and food calorie intake.
Regression analysis of food frequency questionnaires indicated
this may be mediated by increased consumption of energy
dense, low nutrient foods commonly advertised on TV but the
design did not provide direct evidence of this.
Two studies found variations in consumption behaviour,
according to exposure to food promotion, but results were not
statistically significant (Jeffrey, McLellarn & Fox, 1982 Study 1;
Dawson et al., 1988). Four studies produced inconclusive
results. In Galst (1980) and Peterson et al. (1984) it was not
possible to disentangle the effects of food promotion from other
experimental stimuli examined at the same time. Two studies
(Cantor, 1981; Gorn & Goldberg, 1980a) found exposure to food
promotion under certain conditions had an effect on consumption
behaviour but under other conditions it did not, and therefore no
overall consistency in direction of the potential effect was
apparent.
Overall, the studies provide modest evidence that food promotion
does influence consumption behaviour under certain conditions.
However, there were some inconsistencies in some studies and
some did not find any statistically positive association.
(v) Does food promotion
influence children's diet
and health status?
Seven cross-sectional studies addressed this question. Four
investigated the relationship between television viewing and
children’s diet (Bolton, 1983; Gracey et al., 1996; Taras et al.,
1989; Coon et al., 2001). The other three studies examined
health-related variables: two examined the relationship between
television viewing and obesity (Dietz & Gortmaker, 1985;
Matheson et al., 2004) and one (Wong et al., 1992) examined
the relationship between television and video viewing and blood
cholesterol levels. One of the studies was higher scoring in terms
of quality, five were medium scoring and one was lower scoring.
All four dietary studies found significant associations, of varying
strength, between television viewing and dietary intake. Bolton's
study (1983), the higher scoring study, found that food
advertising exposure as calculated from children’s television
viewing diaries was significantly related to children’s snacking
frequency and lower nutrient efficiency of the diet. Coon et al.
(2001) found a significant association between the television
being on during meals and a poorer quality diet. Others found
weak evidence of a relationship between television watching and
fat intake (Taras et al., 1989; Gracey et al., 1996).
Dietz & Gortmaker (1985) found a significant relationship
between television viewing and obesity. Wong et al. (1992) found
a statistically significant relationship between television
viewing/video game playing and high blood cholesterol levels.
31
Overall, there was evidence of small but significant correlation
between television viewing and diet quality (four studies), obesity
(two studies) and blood cholesterol levels (one study). In six of
the studies, the potential effect of food advertising on this
relationship could not be disentangled from the general effect of
television viewing. The effects may have been attributable to the
impact of the advertising seen while watching television, the
impact of other messages seen while watching television, such
as programme content, or to the sedentary nature of the activity
itself (Dietz & Gortmaker, 1985). Alternatively, it is possible that a
high level of television viewing acts as a marker for a complex
set of attitudes and behaviours within the family, which taken
together lead to observed associations between television and
children’s food-related behaviour and diets (Coon et al., 2001).
However, the Bolton study (1983) did measure the specific
contribution of food advertising. Detailed television viewing
diaries enabled a calculation of the extent to which each subject
was exposed specifically to food advertising rather than simply
the amount of time the subject spent watching television in
general. The results showed that the greater a child’s food
advertising exposure, the more frequent his or her snacking and
the lower the nutrient efficiency of his/her diet.
4.3.3 CONCLUSION
Evidence from simpler, descriptive studies demonstrates that
children in both the developed and developing world have
extensive recall of food advertising. Food adverts are among
their favourite types of advertising, with the most popular being
for chocolate, sweets, soft drinks and other foods high in fat,
sugar and salt (e.g. snacks, fizzy drinks, and chocolate). Their
response to food promotion is carried forward into their
communication and shopping behaviours.
Evidence from many of the more complex studies, capable of
inferring causality, demonstrate a statistically significant
association between food promotion and children’s knowledge,
attitudes, behaviours and health status. In terms of nutrition
knowledge, food advertising seems to have little influence on
children’s general perceptions of what constitutes a healthy diet,
but in certain contexts it does have an effect on more specific
nutrition knowledge. There is also evidence that food promotion
influences children’s food preferences, and encourages
purchase and purchase requests to parents for the advertised
foods. Food promotion was also found to influence children’s
consumption and other diet-related outcomes. These effects are
significant, independent of other influences and operate at both
brand and category level.
None of the more complex studies were undertaken in the
developing world. However, surveys from countries in Asia, the
Middle East and South America demonstrate that children are
influenced to try advertised foods and often request parents to
buy advertised foods. Parents – especially those from
disadvantaged backgrounds – frequently accede to these
requests. Disadvantaged mothers also attribute more importance
and credibility to the advertised food products and the advertising
messages, than their more privileged peers. Overall children in
developed and developing countries are found to respond to food
promotion in similar ways. It is therefore highly likely that any
evidence of cause and effect from developed countries is
32
applicable to children in developing and emerging economies
also.
4.4 What is the
extent of food
promotion's
influence relative
to other factors?
Eight cross-sectional studies investigated the relative magnitude
of effect of food promotion or television viewing compared to
other potential influential factors on children’s dietary status.
Research clearly indicates that socioeconomic status is
associated with dietary behaviours and outcomes. However, in
most of the studies reviewed here, this is treated as a
confounding variable and is therefore not discussed here.
Bolton (1983) and Ritchey & Olson (1983) explored the relative
influence of parental dietary behaviours and determinants of
behaviour on children’s behaviours, determinants of behaviour
and diet-related outcomes. Bolton (1983) found that parental
snacking behaviour, and energy intake and nutrient efficiency of
parental diet were significantly associated with children’s own
dietary behaviours. The authors do not provide all of the data on
relative effects, but do note that “the combined direct and indirect
effect of food commercial exposure on children’s calorific intake
and nutrient efficiency was at most half the size of the direct
impact of other predictor variables”. Ritchey & Olson (1983)
found parents’ own diet preferences and consumptions, along
with TV viewing, was significantly associated with children’s selfreported/parental-reported
behaviours
and
preferences,
explaining 35% of the variance in children’s preferences. In this
study also, although TV viewing was associated with children’s
preferences and other related variables, exact estimates of the
contribution TV viewing to the variance of children’s preferences
was not presented.
Associations between parental supervision and control of
children’s exposure to food advertising and dietary intake and
dietary outcomes were investigated by Bolton (1983), Buijzen,
Schuurman & Bomhof (2008) and Norton, Falciglia & Ricketts
(2000). Bolton (1983) did not find the association between
parental supervision and diet to be statistically significant.
However, Buijzen, Schuurman & Bomhof (2008) found
advertising
exposure
and
consumption-related
family
communication styles were significantly associated with
consumption of energy dense products and advertised products.
Correlation coefficients were stronger between advertising and
food consumption than between communication styles and
consumption. The reporting of the data analysis confirms the
interaction effects, but does not allow the relative influences to
be fully disentangled quantitatively. Norton, Falciglia & Ricketts
(2000) found parental food provision was significantly associated
with preferences but this was less influential than taste, exposure
to advertising, and peer behaviours. Again, no standardized
regression coefficients were reported so direct quantitative
comparison is not possible.
The influence of peers and friends was investigated by Dietz &
Gortmaker (1985) and Norton, Falciglia & Ricketts (2000). Dietz
& Gortmaker (1985) found no significant association between
friendship and weight status, but Norton, Falciglia & Ricketts
(2000) found ‘peers eat it’ was significantly correlated with
preference but the magnitude of effect was slightly less than
33
advertising exposure.
Overall, parental and peer influence appears to have small but
significant effect in some circumstances.
Television watching was treated as both a proxy measure of
exposure to TV advertising and an independent variable in its
own right. Clearly, TV viewing has the potential to influence
outcomes in many ways, for example as sedentary activity
replacing more active lifestyles, as an activity associated with
snacking or ‘grazing’, as a communication channel for other
messages in addition to advertising and so on. Results of studies
reviewed are mixed but on the whole find TV viewing correlates
with less desirable outcomes such as higher dietary fat intake.
Wong et al. (1992) found that time spent watching television and
playing video games was a significant and independent predictor
of raised cholesterol in children. Dietz & Gortmaker (1985) found
that TV viewing was predictive, at marginally significant levels, of
obesity and that this effect occurred independently of prior
obesity and family socioeconomic characteristics. Coon et al.
(2001) found that the television being on during meals had a
significant and independent influence on children’s diet.
However, it was not possible, from the results presented, to
judge the strength of the influence of television during meals
relative to the other influences examined. Norton, Falciglia &
Ricketts (2000) found that television advertising was significantly
associated with preferences for a small number of foods, and
that this occurred independently of other motivational factors
influencing food preferences. A study by Gracey et al. (1996)
provided weak evidence that television watching had a small,
marginally significant, independent influence on fat intake and
nutrition knowledge, but the relative strength of the influence of
television watching was not assessed. In summary, television
viewing appears to be independently associated with a number
of indicators of less optimal diet, dietary determinants and
outcomes, in addition to its influences as a main source of food
promotion exposure.
Many of the studies reviewed do not provide sufficient data to
allow a quantitative assessment of the relative influences.
However, the evidence consistently indicates that advertising
and other forms of food promotion are significantly influencers of
children’s food behaviours and dietary outcomes on a magnitude
either similar to or greater than other effects investigated.
4.5 Does food
promotion affect
total category
sales, brand
switching or both
Six studies examined the relationship between food promotion,
brand preferences, and behaviour. Nine studies examined
category preferences and behaviour. The latter group of studies
looked specifically at whether food promotion caused children to
prefer or consume more foods in ‘less healthy’ categories than
foods in a ‘more healthy’ categories. All the studies were
conducted in North America, except one conducted in Europe.
Five were higher scoring in terms of quality, nine were medium
scoring, and one was lower scoring.
Two of the brand preference studies (Gorn & Goldberg, 1980a;
Borzekowski & Robinson, 2001) found that exposure to food
promotion significantly increased children’s likelihood of selecting
34
the advertised food over a non-advertised food. Two studies
found that promotion had no effect on brand preferences (Clarke,
1984; Gorn & Florsheim, 1985), and one found only very modest
effects in favour of the advertised brand (Heslop & Ryans, 1980).
Robinson et al. (2007) found children aged 3-5 years preferred
the taste of foods and drinks presented as McDonald’s branded
products over identical products presented in plain, unbranded
packaging. This effect applied both to products sold by
McDonald’s and products (carrots) not sold by the restaurant
chain at the time of the experiment, indicating brand may be a
powerful motivational variable across many food categories, and
beyond those featured in promotion and advertising.
Of the nine studies which compared children’s preferences or
behaviour in relation to foods in higher fat, sugar or salt
categories versus foods in lower fat, sugar or salt categories,
four found that they were more likely to select higher fat, sugar or
salt products in a one-off preferences test (Goldberg, Gorn &
Gibson, 1978a/1978b Study 1; Stoneman & Brody, 1981;
Kaufman & Sandman, 1983) or for a daily snack (Gorn &
Goldberg, 1980b/1982). Goldberg, Gorn & Gibson, 1978a/1978b
Study 2 found no significant effects on category preferences, but
Buijzen, Schuurman & Bomhof (2008) found brand and category
preference around promoted food products. Buijzen, Schuurman
& Bomhof (2008) also reported that the association between
exposure and consumption of branded products was stronger for
higher income families, and in lower income families the strength
of association shifted to category consumption and exposure to
advertising. The researchers hypothesized this may reflect the
greater purchasing power of higher income families to respond to
specific brand purchase requests, but this pathway was not
explicitly tested. Galst (1980), Peterson et al. (1984) and Cantor
(1981) investigated the association between promotion and food
category preference and consumption, but overall produced
inconclusive results: Cantor (1981) found exposure to food
promotion under certain conditions increased children’s tendency
to consume more dessert foods from a ‘sweet’ category rather
than fruit, but under other conditions the effect became nonsignificant. In addition, Gorn & Florsheim (1985) found no effect
on category preferences.
Although no study provides a thorough comparison of the
strength of both types of effect, both types of effect have been
examined independently, and there is reasonably strong
evidence that both occur. The balance of evidence therefore
suggests that the effects of food promotion on food behaviours
occur at category level as well as encouraging brand switching.
35
5. DISCUSSION
This review updates the 2003, 2004 and 2006 reviews (Hastings
et al., 2003; McDermott et al., 2004; Hastings et al., 2006). It
confirms that in both developed and developing countries: (i)
there is a great deal of food promotion to children, particularly in
the form of television advertising; (ii) this is typically for highly
processed, energy dense, unhealthy products with strong,
evocative branding; and (iii) that children recall, enjoy and
engage with this advertising.
Children across the world therefore, are exposed to marketing
that promotes an unhealthy diet, based on high fat, salt and
sugar, energy-dense, low micronutrient products. Globalization of
trade, and specifically the opening up of massive markets in
India, South America and China, enables massive increases in
the reach and impact of this marketing activity.
Establishing the effects of all this promotional activity requires
complex, resource-intensive research. The evidence base is not
complete and perhaps this will always be the case. However, the
evidence to date does clearly illustrate that food promotion does
influence children’s food preferences, purchase behaviour and
consumption, and that these effects are significant, independent
of other influences and operate at both brand and category level.
The evidence is consistent and continues to amass that the
marketing trends in the developed world are increasingly used in
the developing world. There is no basis on which to infer that
children in the developing world will be any less vulnerable to
food promotion reach and effects. In fact, the evidence available
confirms that they are responsive to promotion and may in fact
spearhead marketing-driven cultural innovation and change.
This review almost certainly underestimates the impact of food
promotion. The evidence base to date still focuses on television
advertising, with relatively less attention given to other forms of
advertising and the full marketing mix. Other forms of promotion
(e.g. merchandising, packaging, sponsorships), product
development, pricing and distribution strategies may each
influence consumer responses independently and also as
integrated, synergistic mechanisms. Measuring this effect
holistically in its real world setting, while adjusting for the myriad
of additional influencing factors is unlikely to ever fully capture
the cumulative effect. Yet the combination of the full marketing
tool box that underpins the most powerful food brands is of
course the real strength of a large scale marketing strategy.
36
REFERENCES
For new studies published between 1st March 2006 and 15th November 2008:
*
indicates those on the extent and nature of food promotion to children
**
indicates those on the effects of food promotion to children
***
indicates those on both the extent and nature, and the effects, of food promotion to
children.
EXTENT AND NATURE REFERENCES
*** Aktaş Arnas Y (2006). The effects of television food advertisement on children's food purchasing
requests. Pediatrics International, 48(2):138-45.
Alexander A et al. (1998). “We’ll be back in a moment”: a content analysis of advertisements in
children’s television in the 1950s. Journal of Advertising, 27(3):1–9.
* Alvy LM, Calvert SL (2008). Food marketing on popular children's web sites: A content analysis.
Journal of the American Dietetic Association, 108(4):710-713.
Atkin CK (1975a). The Effects of Television Advertising on Children. Report No 5: Content Analysis of
Children’s Television Commercials. East Lansing, MI: Michigan State University.
Atkin CK, Heald G (1977). The content of children’s toy and food commercials. Journal of
Communication, 27(1):107–114.
Bang H-K, Reece BB (2003). Minorities in children’s television commercials: New, improved, and
stereotyped. The Journal of Consumer Affairs, 37(1):42–67.
Barcus FE (1971a). Description of Children’s Television Advertising. Boston, MA: School of Public
Communication, Boston University.
Barcus FE (1971b). Saturday Children’s Television: A Report of TV Programming and Advertising on
Boston Commercial Television. Newtonville, MA, Action for Children’s Television.
Barcus FE (1975a). Television in the Afterschool Hours: A Study of Programming of Advertising for
Children on Independent Stations across the United States. Boston, MA, Action for Children’s
Television.
Barcus FE (1975b). Weekend Commercial Children’s Television: A study of programming and
advertising to children on five Boston stations. Newtonville, MA, Action for Children’s Television.
Barcus FE (1981). The nature of television advertising to children. In: Palmer EL, Dorr A, eds.
Children and the Faces of Television: Teaching, Violence and Selling. New York, NY, Academic
Press, 273–285.
Barcus FE, Wolkin R (1977). Children’s Television: an Analysis of Programming and Advertising. New
York, NY, Praeger Publishers.
37
*** Barnabè et al. (2008). The Effect of Advertising and Marketing Practices on Child Obesity. For the
European Parliament's Committee on the Environment, Public Health and Food Safety
(IP/A/ENVI/ST/2007-16). Policy Department, Economic and Scientific Policy, European Parliament,
February.
PE
393.525
(http://www.europarl.europa.eu/activities/committees/studiesCom/download.do?file=19308, accessed
10 November 2008).
* Batada A, Wootan MG (2007). Nickelodeon markets nutrition-poor foods to children. American
Journal of Preventive Medicine, 33(1):48-50.
* Batada A et al. (2008). Nine out of 10 food advertisements shown during Saturday morning
children's television programming are for foods high in fat, sodium, or added sugars, or low in
nutrients. Journal of the American Dietetic Association, 108(4):673-678.
* Berry B, McMullen T (2008). Visual communication to children in the supermarket context: Health
protective or exploitive? Agriculture and Human Values, 25(3):333-348.
Buijzen M, Valkenburg PM (2002). Appeals in television advertising: A content analysis of
commercials aimed at children and teenagers. Communications, 27(3):349–364.
Byrd-Bredbenner C (2002). Saturday morning children’s television advertising: A longitudinal content
analysis. Family and Consumer Sciences Research Journal, 30(3):382–403.
Byrd-Bredbenner C, Grasso D (1999a). A comparative analysis of television food advertisements and
current dietary recommendation. American Journal of Health Studies, 15(4):169–180.
Byrd-Bredbenner C, Grasso D (1999b). Prime-time health: an analysis of health content in television
commercials broadcast during programs viewed heavily by children. International Electronic Journal
of Health Education, 2(4):159–169.
Byrd-Bredbenner C, Grasso D (2000a). Health, medicine, and food messages in television
commercials during 1992 and 1998. The Journal of School Health, 70(2):61–65.
Byrd-Bredbenner C, Grasso D (2000b). Trends in US prime-time television food advertising across
three decades. Nutrition and Food Science, 30(2):59–66.
Byrd-Bredbenner C, Grasso D (2000c). What is television trying to make children swallow?: Content
analysis of the nutrition information in prime-time advertisements. Journal of Nutrition Education,
32(4):187–195.
*** Carter OB (2006). The weighty issue of Australian television food advertising and childhood
obesity. Health Promotion Journal of Australia, 17(1):5-11.
* Chapman K, Nicholas P, Supramaniam R (2006). How much food advertising is there on Australian
television? Health Promotion International, 21(3):172-180.
* Chapman K et al. (2006). The extent and nature of food promotion directed to children in Australian
supermarkets. Health Promotion International, 21(4):331-339.
* Chester J, Montgomery K (2007). Interactive Food & Beverage Marketing: Targeting Children and
Youth in the Digital Age. Berkeley, CA, Berkeley Media Studies Group, The Public Health Institute.
* Chester J, Montgomery K (2008). Interactive Food & Beverage Marketing: Targeting Children and
Youth in the Digital Age. An Update. Berkeley, CA, Berkeley Media Studies Group, The Public Health
Institute.
Chestnutt IG, Ashraf FJ (2002). Television advertising of foodstuffs potentially detrimental to oral
health: A content analysis and comparison of children’s and primetime broadcasts. Community Dental
Health, 19(2):86–89.
Childs NM, Maher JK (2003). Gender in food advertising to children: Boys eat first. British Food
Journal, 105(7):408–419.
Choate RB (1972). The sugar coated children’s hour. The Nation, 31 January 1972:146–148.
Condry JC, Bence PJ, Scheibe CL (1988). Nonprogram content of children’s television. Journal of
Broadcasting and Electronic Media, 32(3):255–270.
* Connor SM (2006). Food-related advertising on preschool television: Building brand recognition in
young viewers. Pediatrics, 118(4):1478-1485.
38
Consumers International (1996). A Spoonful of Sugar – Television food advertising aimed at children:
An international comparative study. London, Consumers International Programme for Developed
Economies.
Consumers International (1999). Easy Targets – A Survey of Television Food and Toy Advertising to
Children in Four Central European Countries. London, Consumers International.
Consumers Union (1995). Captive Kids: A report on commercial pressures on kids at school. Guide
prepared for “ZILLIONS: For Kids” from Consumer Reports. Yonkers, NY, Consumers Union
Education Services, Consumers Union of US, Inc. (http://www.consumersunion.org/other/captivekids,
accessed 10 April 2003).
Cotugna N (1988). TV ads on Saturday morning children’s programming – what’s new? Journal of
Nutrition Education, 20(3):125–127.
* Cowburn G, Boxer A (2007). Magazines for children and young people and the links to Internet food
marketing: a review of the extent and type of food advertising. Public Health Nutrition, 10(10):10241031.
CWS Ltd. (2000). Blackmail – The first in a series of inquiries into consumer concerns about the
ethics of modern food production and advertising. Manchester, Co-operative Group (CWS) Ltd.
Dibb SE (1993). Children: Advertiser’s Dream, Nutrition Nightmare? NFA, London.
Dibb S, Castell A (1995). Easy to swallow, hard to stomach: The results of a survey of food
advertising on television. London, National Food Alliance.
Dibb S, Gordon S (2001). TV Dinners: What’s being served up by the advertisers? London, Sustain The Alliance for Better Food and Farming.
Dickinson R (1997). Television and Food Choice. London, Ministry of Agriculture, Fisheries and Food
(MAFF).
Dickinson R (2000). Food and eating on television: Impacts and influences. Nutrition and Food
Science, 30(1):24–29.
* Dixon H, Scully M, Parkinson K (2006). Pester power: snackfoods displayed at supermarket
checkouts in Melbourne, Australia. Health Promotion Journal of Australia, 17(2):124-127.
Doolittle J, Pepper R (1975). Children’s TV Ad Content: 1974. Journal of Broadcasting, 19:131–142.
* Eagle L, Brennan R (2007). Beyond advertising: In-home promotion of ‘fast food’. Young
Consumers, 8(4):278-288.
Egberts K, Riley M (2004). Food advertisements during children’s and adult’s viewing times: A
comparative study. Asia Pacific Journal of Clinical Nutrition, 13(Supplement):S60.
* Elliott C (2008a). Marketing fun foods: A profile and analysis of supermarket food messages
targeted at children. Canadian Public Policy-Analyse De Politiques, 34(2):259-273.
* Elliott C (2008b). Assessing ‘fun foods’: nutritional content and analysis of supermarket foods
targeted at children. Obesity Reviews, 9(4):368-377.
* Federal Trade Commission (2008). Marketing Food to Children and Adolescents. A Review of
Industry Expenditures, Activities, and Self-Regulation. A Report to Congress. Washington, DC,
Federal Trade Commission. (Appendices published as separately: Federal Trade Commission
(2008). Marketing Food to Children and Adolescents. A Review of Industry Expenditures, Activities,
and Self-Regulation. A Report to Congress. Appendices. Washington, DC, Federal Trade
Commission.)
* Folta SC et al. (2006). Food advertising targeted at school-age children: A content analysis. Journal
of Nutrition Education and Behavior, 38(4):244-248.
* Galcheva SV, Iotova VM, Stratev VK (2008). Television food advertising directed towards Bulgarian
children. Archives of Disease in Childhood, 93(10):857-861.
Gamble M, Cotugna N (1999). A quarter century of TV food advertising targeted at children. American
Journal of Health Behavior, 23(4):261–267.
39
* Gantz W et al. (2007). Food for Thought: Television Food Advertising to Children in the United
States. Kaiser Family Foundation Report #7618. Menlo Park, CA, Kaiser Family Foundation.
*** Guittard C (2008). Note 1 From Ms Christine Guittard. In: Guittard C & Sjölin K. Advertising and
marketing practices on child obesity. Compilation. For the European Parliament's Committee on the
Environment, Public Health and Food Safety (IP/A/ENVI/NT/2007-20 & 21). Policy Department,
Economic and Scientific Policy, European Parliament, February. PE 400.989. (http://www.pedz.unimannheim.de/daten/edz-ma/ep/08/EST19988.pdf, accessed 17 November 2008).
Gussow J (1972). Counternutritional messages of TV ads aimed at children. Journal of Nutrition
Education, 4:48–52.
Gussow J (1973). It makes even milk a dessert – a report on the counternutritional messages of
children’s television advertising. Clinical Paediatrics, 12:68–71.
Hammond K, Wyllie A, Casswell S (1997a). Content analysis of television food advertisements.
Auckland, Alcohol & Public Health Research Unit.
Hammond K, Wyllie A, Casswell S (1997b). The extent and nature of televised food advertising in
New Zealand. Auckland, Alcohol & Public Health Research Unit.
Hammond KM, Wyllie A, Casswell S (1999). The extent and nature of televised food advertising to
New Zealand children and adolescents. Australian and New Zealand Journal of Public Health,
23(1):49–55.
* Harrison K (2006). Fast and Sweet: Nutritional Attributes of Television Food Advertisements With
and Without Black Characters. The Howard Journal of Communications, 17(4):249-264.
Harrison K, Marske AL (2005). Nutritional content of foods advertised during the television programs
children watch most. American Journal of Public Health, 95(9):1568–1574.
Hawkes C (2002). Marketing activities of global soft drink and fast food companies in emerging
markets: A review. In: Globalization, Diets and Noncommunicable Diseases. Geneva, World Health
Organization.
Hill JM, Radimer KL (1997). A content analysis of food advertisements in television for Australian
children. Australian Journal of Nutrition and Dietetics, 54(4):174.
* Ho CC, Len YK (2008). Cereal Deceptors. The Marketing of Breakfast Cereals to Children in
Malaysia. Consumers International Junk Food Generation Campaign. Selangor, The Federation of
Malaysian Consumers Association (FOMCA).
* Holt DJ et al. (2007). Children's Exposure to TV Advertising in 1977 and 2004: Information for the
Obesity Debate. Washington, DC, Bureau of Economics, Federal Trade Commission.
Horgen KB, Choate M, Brownell KD (2001). Television Food Advertising. Targeting children in a toxic
environment. In: Singer DG, Singer JL, eds. Handbook of Children and the Media. Thousand Oaks,
CA, Sage Publications Inc, Chapter 22.
Howard SR (2003). Innocent little 30-second tales: How children’s food commercials normalize social
alienation, violence, crime, and substance use. A content analysis of children’s food commercials,
1987-1998. Dissertation Abstracts International: Section B: The Sciences and Engineering, 63(10B):4964.
Ji MF, McNeal JU (2001). How Chinese children’s commercials differ from those of the United States:
A content analysis. Journal of Advertising, 30(3):79–92.
* Jones S, Fabrianesi B (2006). Gross for kids but good for parents: differing messages in
nd
advertisements for the same products. 2 bi-annual “Social Marketing Advances in Research and
Theory”
(SMART)
conference,
Banff,
Canada,
19-21
October,
2006
(http://www.uleth.ca/man/research/centres/csrm/conferences/smart/2006/papers/09.pdf, accessed 18
November 2008).
* Jones S, Fabrianesi B (2008). Gross for kids but good for parents: differing messages in
advertisements for the same products. Public Health Nutrition, 11(6):588-595.
* Karupaiah T et al. (2008). What's on Malaysian television? A survey on food advertising targeting
children. Asia Pacific Journal of Clinical Nutrition, 17(3):483-491.
40
* Kelly B, Chapman K (2007). Food references and marketing to children in Australian magazines: a
content analysis. Health Promotion International, 22(4):284-291.
* Kelly B et al. (2007). Television food advertising to children: the extent and nature of exposure.
Public Health Nutrition, 10(11):1234-1240.
* Kelly B et al. (2008a). Internet food marketing on popular children's websites and food product
websites in Australia. Public Health Nutrition, 11(11):1180-1187.
* Kelly B et al. (2008b). Persuasive food marketing to children: cartoons and competitions on Sydney
commercial television. International Journal of Obesity, 32(Suppl1):S182.
* Kelly B et al. (2008c). Persuasive food marketing to children: use of cartoons and competitions in
Australian commercial television advertisements. Health Promotion International, 23(4):337-344.
Klebba JM, Stern BL, Tseng D (1994). Disclaimers in children’s television advertising revisited: a
decade makes a difference. Proceedings of the 1994 Conference of the American Academy of
Advertising, 50–57.
Kotz K, Story M (1994). Food advertisements during children’s Saturday morning television
programming - are they consistent with dietary recommendations. Journal of the American Dietetic
Association, 94(11):1296–1300.
Kunkel D, Gantz W (1992). Children’s television advertising in the multichannel environment. Journal
of Communication, 42(3):134–152.
Kuribayashi A, Roberts MC, Johnson RJ (2001). Actual nutritional information of products advertised
to children and adults on Saturday. Children’s Health Care, 30(4):309–322.
Larson MS (2003). Gender, race and aggression in television commercials that feature children. Sex
Roles, 48(1/2):67–75.
Lemos DMP (2004). Saturday morning children television food advertising … The nightmare of
nutrition educators. Journal of Pediatric Gastroenterology and Nutrition, 39(Suppl 1):S471–S472.
Lewis MK, Hill AJ (1998). Food advertising on British children’s television: A content analysis and
experimental study with nine-year olds. International Journal of Obesity, 22(3):206–214.
* Lobstein T et al. (2008). Cereal Offences: A wake-up call on the marketing of unhealthy food to
children. London, Consumers International.
Longman B (2000). Marketing Food & Drinks to Kids: Driving Appeal to Young Consumers. London,
Reuters Business Insight, Consumer Goods.
Longman B (2002). Marketing Food & Drinks to the Internet Generation: Effective youth targeting and
profit opportunities. London, Reuters Business Insight, Consumer Goods.
Macklin MC, Kolbe RH (1984). Sex role stereotyping in children’s advertising: current and past trends.
Journal of Advertising, 13(2):34–42.
* Maher JK et al. (2006). Food advertising on children's television. Young Consumers, 7(4):41-52.
Maryam A et al. (2005). Food advertising on Iranian children’s television: A content analysis and an
experimental study with junior high school students. Ecology of Food and Nutrition, 44(2):123–133.
* Matthews AE (2008). Children and obesity: a pan-European project examining the role of food
marketing. European Journal of Public Health, 18(1):7-11.
* Mazur A et al. (2008). Impact of food advertising on food primary and secondary schools in
purchases by students in south-eastern Poland. Public Health Nutrition, 11(9):978-981.
Messner M et al. (1999). Boys to Men: Sports Media. Messages about Masculinity: A National Poll of
Children, Focus Groups, and Content Analysis of Sports Programs and Commercials. Oakland, CA,
Children Now Research.
* Molnar A et al. (2008). Marketing of foods of minimal nutritional value to children in schools.
Preventive Medicine, 47(5):504-507.
* Moore ES (2006). It’s Child’s Play: Advergaming and the Online Marketing of Food to Children.
Kaiser Family Foundation Report # 7536. Menlo Park, CA, Kaiser Family Foundation.
41
* Moore ES (2008). Food marketing goes online: A content analysis of Web sites for children. In:
Fitzgerald HE, Mousouli V, Davies HD, eds. Obesity in childhood and adolescence, Vol 2:
Understanding development and prevention, 93-115.
* Moore ES, Rideout VJ (2007). The Online marketing of food to children: Is it just fun and games?
Journal of Public Policy & Marketing, 26(2):202-220.
* Morgan M et al. (2008). A content analysis of children's television advertising: focus on food and
oral health. Public Health Nutrition, [Epub]:1-8.
Morton H (1984). The television advertising of foods to children - A south Australian study. Journal of
Food and Nutrition, 41(4):170–175.
Morton H (1990). Television food advertising - A challenge for the new public health in Australia.
Community Health Studies, 14(2):153–161.
Morton H, Moore V (2004). Childhood obesity, declining food skills and television food advertising.
International Journal of Obesity, 28(Suppl 3):S128.
Muehling DD, Kolbe RH (1998). A comparison of children’s and prime-time fine-print advertising
disclosure practices. Journal of Advertising, 27(3):37–48.
Neville L, Thomas M, Bauman A (2005). Food advertising on Australian television: The extent of
children's exposure. Health Promotion International, 20(2):105–112.
Ogletree SM et al. (1990). Female attractiveness and eating disorders: Do children’s television
commercials play a role? Sex Roles, 22(11–12):791–797.
* Outley CW, Taddese A (2006). A content analysis of health and physical activity messages
marketed to African American children during after-school television programming. Archives of
Pediatrics & Adolescent Medicine, 160(4):432-435.
* Page RM, Brewster A (2007a). Frequency of promotional strategies and attention elements in
children's food commercials during children's programming blocks on US broadcast networks. Young
Consumers, 8(3):184-196.
* Page RM, Brewster A (2007b). Emotional and rational product appeals in televised food
advertisements for children: Analysis of commercials shown on US broadcast networks. Journal of
Child Health Care, 11(4):323-340.
* Powell LM, Szczypka G, Chaloupka FJ (2007a). Exposure to food advertising on television among
US children. Archives of Pediatrics & Adolescent Medicine, 161(6):553-560.
* Powell LM, Szczypka G, Chaloupka FJ (2007b). Adolescent exposure to food advertising on
television. American Journal of Preventive Medicine, 33(4):S251-S256.
* Powell LM et al. (2007). Nutritional content of television food advertisements seen by children and
adolescents in the United States. Pediatrics, 120(3):576-583.
Rajecki DW, McTavish DG, Jessup KS (1994). Violence, conflict, trickery, and other story themes in
TV ads for food for children. Journal of Applied Social Psychology, 24(19):1685–1700.
Reece BB, Rifon NJ, Rodriguez K (1999). Selling Food to Children: Is fun part of a balanced
breakfast? In: Macklin MC, Carlson L, eds. Advertising to Children Concepts and Controversies,
Thousand Oaks, CA, Sage Publications Ltd., Chapter 9.
* Roberts M, Pettigrew S (2007). A thematic content analysis of children's food advertising.
International Journal of Advertising, 26(3):357-367.
* Robinson E (2008). The Junk Food Trap. Marketing Unhealthy Food to Children in Asia Pacific.
London, Consumers International.
Rodd HD, Patel V (2005). Content analysis of children’s television advertising in relation to dental
health. British Dental Journal, 199(11):710–712.
* Schwartz MB et al. (2008). Examining the nutritional quality of breakfast cereals marketed to
children. Journal of the American Dietetic Association, 108(4):702-5. [Comments in: Journal of the
American Dietetic Association, 2008 Oct;108(10):1616-7; author reply 1619-20.]
42
*** Sjölin K (2008). Note 2 From Ms Kristina Sjölin. In: Guittard C & Sjölin. Advertising and marketing
practices on child obesity. Compilation. For the European Parliament's Committee on the
Environment, Public Health and Food Safety (IP/A/ENVI/NT/2007-20 & 21). Policy Department,
Economic and Scientific Policy, European Parliament, February. PE 400.989 (http://www.pedz.unimannheim.de/daten/edz-ma/ep/08/EST19988.pdf, accessed 17 November 2008).
Solomon H et al. (1982). Content and Effect of Children’s Commercials. Paper presented at the
Annual Meeting of the Eastern Psychological Association, Baltimore, April 1982. Available from the
Educational Resources Information Center, Institute of Education Sciences, US Department of
Education; clearinghouse number: IR010579.
Stern BL, Harmon RR (1984). The incidence and characteristics of disclaimers in children’s television
advertising. Journal of Advertising, 13(2):12–16.
Taras H et al. (2000). Television advertising and classes of food products consumed in a paediatric
population. International Journal of Advertising, 19(4):487–494.
Taras HL, Gage M (1995). Advertised foods on children’s television. Archives of Pediatrics and
Adolescent Medicine, 149(6):649–652.
* Temple NJ, Steyn NP & Nadomane Z (2008). Food advertisements on children's programs on TV in
South Africa. Nutrion, 24(7-8):781-782.
* Thompson DA et al. (2008). Comida en venta: After-school advertising on Spanish-language
television in the United States. Journal of Pediatrics, 152(4):576-581.
* Warren R et al. (2007). Food and beverage advertising to children on US television: Did national
food advertisers respond? Journalism & Mass Communication Quarterly, 84(4):795-810.
* Warren R et al. (2008). Food and beverage advertising on US television: A comparison of childtargeted versus general audience commercials. Journal of Broadcasting & Electronic Media,
52(2):231-246.
* Weber K, Story M, Harnack L (2006). Internet food marketing strategies aimed at children and
adolescents: A content analysis of food and beverage brand Web sites. Journal of the American
Dietetic Association, 106(9):1463-1466.
* Which? (2008). Food Fables – the second sitting. The truth behind how food companies target
children. London, Which? (Consumers’ Association).
Wilson N, Quigley R, Mansoor O (1999). Food ads on TV: A health hazard for children? Australian
and New Zealand Journal of Public Health, 23(6):647–650.
Wilson N et al. (2006a). Marketing fat and sugar to children on New Zealand television. Preventive
Medicine, 42(2):96–101.
Wilson N et al. (2006b). Hazardous and beneficial nutritional messages in 858 televised food
advertisements during children’s viewing hours. New Zealand Medical Journal, 119(1233):U1967.
Winick C et al. (1973). Children’s Television Commercials – A Content Analysis. New York, Praeger
Publishers.
Young B (1987). Sugar, children and television advertising. Research Report No, 15. London, Health
Education Authority.
Young B (1990). What Children Watch: American and British television advertising compared. In:
Young B, Television Advertising and Children. Oxford, Clarendon Press, Chapter 6.
Young B et al. (1996). The Role of Advertising in Children’s Food Choice. London, Ministry of
Agriculture, Foods and Fisheries (MAFF).
Zuppa JA, Morton H, Mehta KP (2003). Television food advertising: Counterproductive to children’s
health? A content analysis using the Australian guide to healthy eating. Nutrition and Dietetics,
60(2):78–84.
43
EFFECTS REFERENCES
*** Aktaş Arnas Y (2006). The effects of television food advertisement on children's food purchasing
requests. Pediatrics International, 48(2):138-45.
Arluk SL et al. (2003). Childhood obesity’s relationship to time spent in sedentary behaviour. Military
Medicine, 168(7):583–586.
Atkin CK (1975b). The Effects of Television Advertising on Children. Report No. 7: Parent-Child
Communication in Supermarket Breakfast Selection. Washington, DC, Office of Child Development
(DHEW).
Atkin CK (1975c). The Effects of Television Advertising on Children. Report No. 6: Survey of PreAdolescent’s Responses to Television Commercials. Washington, DC, Office of Child Development
(DHEW).
Atkin CK (1978). Observation of parent-child interaction in supermarket decision-making. Journal of
Marketing, 42(4):41–45.
*** Barnabè D et al. (2008). The Effect of Advertising and Marketing Practices on Child Obesity. For
the European Parliament's Committee on the Environment, Public Health and Food Safety
(IP/A/ENVI/ST/2007-16). Policy Department, Economic and Scientific Policy, European Parliament,
February.
PE
393.525
(http://www.europarl.europa.eu/activities/committees/studiesCom/download.do?file=19308, accessed
10 November 2008).
Barry TE, Hansen RW (1973). How race affects children’s TV commercials. Journal of Advertising
Research, 13(5):63–67.
** Batada A, Borzekowski D (2008). Snap! Crackle! What? Recognition of cereal advertisements and
understanding of commercials' persuasive intent among urban, minority children in the US. Journal of
Children and Media, 2(1):19-36.
Bolton RN (1983). Modeling the impact of television food advertising on children’s diets. In: Leigh JH,
Martin Jr CR, eds. Current Issues and Research in Advertising. Ann Arbor, MI, Division of Research,
Graduate School of Business Administration, University of Michigan, 173–199.
Borzekowski DLG, Robinson TN (2001). The 30-second effect: An experiment revealing the impact of
television commercials on food preferences of preschoolers. Journal of the American Dietetic
Association, 101(1):42–46.
Boynton-Jarrett R et al. (2003). Impact of television viewing patterns on fruit and vegetable
consumption among adolescents. Pediatrics, 112(6):1321–1326.
** Bridges E, Briesch RA (2006). The 'nag factor' and children's product categories. International
Journal of Advertising, 25(2):157-187.
** Buijzen M, Schuurman J, Bomhof E (2008). Associations between children's television advertising
exposure and their food consumption patterns: A household diary-survey study. Appetite, 50(23):231-239. (See also: Buijzen M, Schuurman J, Bomhof E (2007). De relatie tussen televisiereclame
voor voedingsmiddelen en consumptiepatronen van kinderen. [The relationship between television
advertising and children's food consumption.] Psychologie & Gezondheid, 35(2):66-75. [Dutch
language]).
Cantor J (1981). Modifying children’s eating habits through television ads: effects of humorous
appeals in a field setting. Journal of Broadcasting, 25(1):37–47.
Carruth BR, Goldberg DL, Skinner JD (1991). Do parents and peers mediate the influence of
television advertising on food-related purchases? Journal of Adolescent Research, 6(2):253–271.
Carruth BR et al. (2000). Preschoolers’ food product choices at a simulated point of purchase and
mothers’ consumer practices. Journal of Nutrition Education, 32(3):146–151.
*** Carter OB (2006). The weighty issue of Australian television food advertising and childhood
obesity. Health Promotion Journal of Australia, 17(1):5-11.
44
** Chamberlain LJ, Wang Y, Robinson TN (2006). Does children's screen time predict requests for
advertised products? Cross-sectional and prospective analyses. Archives of Pediatrics & Adolescent
Medicine, 160(4):363-368.
Chan K (2000). Hong Kong children’s understanding of television advertising. Journal of Marketing
Communications, 6:37–52.
Chan K, McNeal JU (2003). Parental concern about television viewing and children’s advertising in
China. International Journal of Public Opinion Research, 15(2):151–166.
** Chernin A (2008). The effects of food marketing on children's preferences: Testing the moderating
roles of age and gender. Annals of the American Academy of Political and Social Science, 615:102118.
Clarke TK (1984). Situational factors affecting preschoolers’ responses to advertising. Academy of
Marketing Science Journal, 12(4):25.
Coon KA et al. (2001). Relationships between use of television during meals and children’s food
consumption patterns. Pediatrics, 107(1):e7 (http://pediatrics.aappublications.org).
Dawson BL et al. (1988). Television food commercials’ effect on children’s resistance to temptation.
Journal of Applied Social Psychology, 18(16, Pt2):1353–1360.
Del Toro W, Greenberg BS (1989). Television commercials and food orientations among teenagers in
Puerto Rico. Hispanic Journal of Behavioral Sciences, 11(2):168–177.
Dickinson R (1997). Television and Food Choice. London, Ministry of Agriculture, Fisheries and Food
(MAFF).
Dietz WH, Gortmaker SL (1985). Do we fatten our children at the television set? Obesity and
television viewing in children and adolescents. Pediatrics, 75(5):807–812.
** Dixon HG et al. (2007). The effects of television advertisements for junk food versus nutritious food
on children's food attitudes and preferences. Social Science & Medicine, 65(7):1311-1323.
Donkin AJM, Neale RJ, Tilston C (1993). Children’s food purchase requests. Appetite, 21(3):291–
294.
Donkin AJM et al. (1992). Children’s food preferences: Television advertising vs nutritional advice.
British Food Journal, 94(9):6–9.
Donohue TR (1975). Effect of commercials on black children. Journal of Advertising Research,
15(6):41–47.
Escalante de Cruz A et al. (2004). The Junk Food Generation. A multi-country survey of the influence
of television advertisements on children. Kuala Lumpur, Malaysia, Consumers International Asia
Pacific Office.
** Folta SC, Bourbeau J, Goldberg JP (2008). Watching children watch food advertisements on TV.
Preventive Medicine, 46(2):177-178.
Fox DT (1981). Children’s television commercials and their nutrition knowledge and eating habits.
PhD Thesis, University of Montana, MT.
French SA et al. (2001). Pricing and promotion effects on low-fat vending snack purchases: The
CHIPS Study. American Journal of Public Health, 91(1):112–117.
Galst JP (1980). Television food commercials and pro-nutritional public service announcements as
determinants of young children’s snack choices. Child Development, 51(3):935–938.
Galst JP, White MA (1976). The unhealthy persuader: the reinforcing value of television and
children’s purchase influencing attempts at the supermarket. Child Development, 47:1089–1096.
Ghani NHA, Zain OM (2004). Malaysian children's attitudes towards television advertising. Young
Consumers, 5(3):41–51.
Goldberg ME (1990). A quasi-experimental assessing the effectiveness of TV advertising directed to
children. Journal of Marketing Research, 27(4):445–454.
Goldberg ME, Gorn GJ, Gibson W (1978a). The effects of TV messages for high and low nutritional
foods on children’s snack and breakfast food choices. Advances in Consumer Research, 5:540–545.
45
Goldberg ME, Gorn GJ, Gibson W (1978b). TV messages for snacks and breakfast foods: do they
influence children’s preferences? Journal of Consumer Research, 5(2):73–81.
Gorn GJ, Florsheim R (1985). The effects of commercials for adult products on children. Journal of
Consumer Research, 11(4):962–967.
Gorn GJ, Goldberg ME (1980a). Children’s responses to repetitive television commercials. Journal of
Consumer Research, 6(4):421–424.
Gorn GJ, Goldberg ME (1980b). TV’s Influence on Children: The Long and the Short of It. Paper
presented at the 88th Annual Convention of the American Psychological Association, Montreal,
Quebec, Canada, September 1–5, 1980.
Gorn GJ, Goldberg ME (1982). Behavioral evidence of the effects of televised food messages on
children. Journal of Consumer Research, 9(2):200–205.
Gracey D et al. (1996). Nutritional knowledge, beliefs and behaviours in teenage school students.
Health Education Research, 11(2):187–204.
*** Guittard C (2008). Note 1 From Ms Christine Guittard. In: Guittard C & Sjölin K. Advertising and
marketing practices on child obesity. Compilation. For the European Parliament's Committee on the
Environment, Public Health and Food Safety (IP/A/ENVI/NT/2007-20 & 21). Policy Department,
Economic and Scientific Policy, European Parliament, February. PE 400.989 (http://www.pedz.unimannheim.de/daten/edz-ma/ep/08/EST19988.pdf, accessed 17 November 2008).
** Halford JC et al. (2007). Beyond-brand effect of television (TV) food advertisements/commercials
on caloric intake and food choice of 5-7-year-old children. Appetite, 49(1):263-7.
** Halford JC et al. (2008a). Children's food preferences: Effects of weight status, food type, branding
and television food advertisements (commercials). International Journal of Pediatric Obesity, 3(1):3138.
** Halford JC et al. (2008b). Beyond-brand effect of television food advertisements on food choice in
children: the effects of weight status. Public Health Nutrition, 11(9):897-904.
Halford JCG et al. (2004). Effect of television advertisements for foods on food consumption in
children. Appetite, 42:221–225.
Harrison K, Marske AL (2005). Nutritional content of foods advertised during the television programs
children watch most. American Journal of Public Health, 95(9):1568–1574.
Heslop LA, Ryans AB (1980). A second look at children and the advertising of premiums. Journal of
Consumer Research, 6(4):414–420.
Hitchings E, Moynihan PJ (1998). The relationship between television food advertisements recalled
and actual foods consumed by children. Journal of Human Nutrition and Dietetics, 11(6):511–517.
Jeffrey DB, McLellarn RW, Fox DT (1982). The development of children’s eating habits: The role of
television commercials. Health Education Quarterly, 9(2-3):174–189.
* Jones S et al. (2007). Food Marketing to Children in Australia. For The Cancer Council Australia’s
Nutrition and Physical Activity Committee Centre for Health Initiatives. Wollongong, University of
Wollongong.
Kaufman L, Sandman PM (1983). Countering children’s sugared food commercials: Do rebuttals
help? Paper at the 34th Annual Meeting of the International Communication Association, San
Francisco, CA, May 24–28, 1984.
Lam PYY (1978). Influence of Television Commercials on Young Children. Master's Thesis, North
Texas State University, TX, May 1978.
Lewis MK, Hill AJ (1998). Food advertising on British children’s television: A content analysis and
experimental study with nine-year olds. International Journal of Obesity, 22(3):206–214.
Lobstein T, Dibb S (2005). Evidence of a possible link between obesogenic food advertising and child
overweight. Obesity Reviews, 6(3):203–208.
** Marshall D, O'Donohoe S, Kline S (2007). Families, food, and pester power: beyond the blame
game? Journal of Consumer Behaviour, 6(4):164-181.
46
Maryam A et al. (2005). Food advertising on Iranian children’s television: A content analysis and an
experimental study with junior high school students. Ecology of Food and Nutrition, 44(2):123–133.
Maskill C et al. (1996). Influences on the eating patterns and food choice of New Zealand Pakeha
adolescents: An overview. Auckland, Alcohol & Public Health Research Unit.
Matheson DM et al. (2004). Children’s food consumption during television viewing. American Journal
of Clinical Nutrition, 79:1088–1094.
Moya De Sifontes ZM, Dehollain PL (1986). Efecto de los medios de comunicacion social en la
adquisicion de alimentos a nivel familiar (The effect of mass communication media on the family’s
food purchasing patterns). Archivos Latinoamericanos de Nutricion, 36(1):166–186.
Musaiger AO et al. (1986/4). Children’s response to television food advertising in Bahrain. Hygie,
V:30–35.
Neeley SM, Schumann DW (2004). Using animated spokes-characters in advertising to young
children. Journal of Advertising, 33(3):7–23.
Norton PA, Falciglia GA, Ricketts C (2000). Motivational determinants of food preferences in
adolescents and pre-adolescents. Ecology of Food and Nutrition, 39(3):169–182.
Olivares S, Yáñez R, Díaz N (2003). Publicidad de alimentos y conductas alimentarias en escolares
de 5° a 8° básico (Food advertising and food behavior in school age children from 5th to 8th grade).
Revista chilena de nutrición, 30(1):36–42.
Olivares SC et al. (1999). Publicidad televisiva y preferencias alimentarias en escolares de la Región
Metropolitana (Television publicity and food preferences of school age children of the Metropolitan
Region). Revista médica de Chile, 127:791-799.
Pavlu D (2001). What mothers think of the children’s television advertising. 41st International Film
Festival for Children and Youth, The Czech Republic, 27 May – 2 June
(http://www.aeforum.org/aeforum.nsf/0/736619d74076102880256a8d0036f010?OpenDocument,
accessed 23 March 2006).
Peterson PE et al. (1984). How pronutrition television programming affects children’s dietary habits.
Developmental Psychology, 20:55–63.
Radkar A, Mundlay A (2001). Impact of TV Food Advertisements on Buying Decisions: Joint project
by Consumers’ International (ROAP) and Mumbai Grahak Panchayat – India. Results of a Survey
from Maharashtra, India. Pune, Centre for Research and Education (http://www.ciroap.org/food,
accessed 4 April 2006).
Reeves B, Atkin CK (1979). The effects of televised advertising on mother-child interactions at the
nd
grocery store. Paper presented at the 62 Annual Meeting of the Association for Education in
Journalism, Houston, TX, August 5–8, 1979.
Riecken R, Yavas U (1990). Children’s general product and brand-specific attitudes towards
television commercials: Implications for public policy and advertising strategy. International Journal of
Advertising, 9(2):136.
Ritchey N, Olson C (1983). Relationships between family variables and children’s preference for
consumption of sweet foods. Ecology of Food and Nutrition, 13:257–266.
** Robinson TN et al. (2007). Effects of fast food branding on young children's taste preferences.
Archives of Pediatrics & Adolescent Medicine, 161(8):792-797.
Ross RP et al. (1980). Children’s television commercials containing nutritional information: When do
they help? When do they hinder? Paper presented at the Biennial Meeting of the Southwestern
Society for Research in Human Development, Lawrence KS, March 27–29, 1980.
Ross RP et al. (1981). Nutritional misinformation of children: A developmental and experimental
analysis of the effects of televised food commercials. Journal of Applied Developmental Psychology,
1(4):329–347.
*** Sjölin K (2008). Note 2 From Ms Kristina Sjölin. In: Guittard C & Sjölin K. Advertising and
marketing practices on child obesity. Compilation. For the European Parliament's Committee on the
Environment, Public Health and Food Safety (IP/A/ENVI/NT/2007-20 & 21). Policy Department,
47
Economic and Scientific Policy, European Parliament, February. PE 400.989 (http://www.pedz.unimannheim.de/daten/edz-ma/ep/08/EST19988.pdf, accessed 17 November 2008).
Stoneman Z, Brody GH (1981). Peers as mediators of television food advertisements aimed at
children. Developmental Psychology, 17(6):853–858.
Stoneman Z, Brody GH (1982). The indirect impact of child-oriented advertisements on mother-child
interactions. Journal of Applied Developmental Psychology, 2:369–376.
Tantivejakul N (2001). The analysis of cultivation effects of television commercials on Thai school
adolescents. Journal of Communication Arts, 19(3):58–67.
Taras H et al. (1989). Television’s influence on children’s diets and physical activity. Journal of
Developmental and Behavioural Paediatrics, 10(4):176–180.
Taras H et al. (2000). Television advertising and classes of food products consumed in a paediatric
population. International Journal of Advertising, 19(4):487–494.
Unnikrishnan N, Bajpai S (1996). Children’s Advertising. In: Unnikrishnan N, Bajpai S, eds. The
Impact of Television Advertising on Children. New Delhi, Sage, 252–277.
** Utter J, Scragg R, Schaaf D (2006). Associations between television viewing and consumption of
commonly advertised foods among New Zealand children and young adolescents. Public Health
Nutrition, 9(5):606-612.
Vaipeyi R (2001). Impact of TV Food Advertisement Parents and Children Survey. India, Voice.
Ward D, Reale G, Levinson D (1972). Children’s perceptions, explanations and judgements of
television advertising: a further exploration. In: Rubinstein EA, Comstock GA, Murray JP, eds.
Television and Social Behavior: Volume IV. Television in Day-to-Day Life; Patterns of Use.
Washington, DC, US Government Printing Office.
** Wiecha JL et al. (2006). When children eat what they watch - Impact of television viewing on
dietary intake in youth. Archives of Pediatrics & Adolescent Medicine, 160(4):436-442.
Williams S (1974). Television and the Young Consumer. An Analysis of Consumer Needs of Children
and a Proposal for the Utilization of Television to Meet These Needs. Presented at the International
Festival of children Television, 31 March – 2nd April, 1974, Kennedy Centre, Washington. San
Francisco, CA, Committee on Children’s Television. Available from the Educational Resources
Information Center, Institute of Education Sciences, US Department of Education; clearinghouse
number: IR000450.
Wiman AR, Newman LM (1989). Television advertising exposure and children’s nutritional
awareness. Journal (Academy of Marketing Science), 17(2):179–188.
Wong ND et al. (1992). Television viewing and pediatric hypercholesterolemia. Pediatrics, 90:75–79.
Yavas U, Abdul-Gader A (1993). Impact of TV commercials on Saudi children’s purchasing
behaviour. Marketing Intelligence and Planning, 11(2):37–43.
OTHER REFERENCES CITED
Abel JD (1978). The Child Audience for Network Television Programming and Advertising. Children’s
Advertising Rulemaking Comment submitted to the Federal Trade Commission, 24 November.
Adler RP et al. (1977). Research on the Effects of Television Advertising on Children. Washington,
DC, National Science Foundation.
Anderson PM, He X (1999). Culture and the fast-food marketing mix in the People’s Republic of
China and the USA: Implications for research and marketing. Journal of International Consumer
Marketing, 11(1):77–95.
Baker B (2000). Consumer preferences in Hong Kong on Western Foods. Hong Kong, US
Agricultural Trade Office, American Consulate General.
48
Beales JH (1978). An Analysis of Exposure to Non-network Television Advertising. Children's
Advertising Rulemaking Comment submitted to the Federal Trade Commission, 21 November.
Bradford-Hill A (1965). The environment and disease: association or causation? Proceedings of the
Royal Society of Medicine, 58:295–300.
Brown J (1976). Graduate students examine TV ads for food. Journal of Nutrition Education,
9(3):120–122.
Castell A (1988). The rattle of a stick in a swill bin. Postgraduate Diploma in Health Education.
Chan K (2005). Store visits and information sources among urban Chinese Children. Journal of
Consumer Marketing, 22(3):178–188.
Children’s Food Campaign (2007). Missing the target. London, Children’s Food Campaign
(http://www.sustainweb.org/pdf/Missing_the_target.pdf, accessed 9 April 2009).
Coalition on Food Advertising to Children (2007). Children’s health or corporate wealth? The case for
banning television food advertising to children. A briefing paper by the Coalition on Food Advertising
nd
to Children (CFAC). 2 edition. January 2007. Woolloomooloo: The Cancer Council New South
Wales (http://www.chdf.org.au/i-cms_file?page=664/CFAC_BriefingPaper_Jan07.pdf, accessed 9
April 2009).
Donkin A (1992). The effects of television advertising on children’s food preferences and the
nutritional implications. PhD thesis, University of Nottingham.
Ekström L, Sandberg H (2007). Söt reklam Och feta ungar. [Sweet advertising and fat kids.] Nordiska
Ministerrådet.
Copenhagen,
Norden.
[Swedish
language,
English
summary].
(http://www.norden.org/pub/velfaerd/konsument/sk/N2007002.pdf, accessed 9 April 2009).
Exchange for Media (2004). Brand Speak, interview with Pankaj Batra, Marketing Director, YUM!
Restaurants
International
(http://www.exchange4media.com/Brandspeak/brandspeak.asp?brand_id=62, accessed 4 April
2006).
Food Commission (1990). Sweet Persuasion – A diet of junk food adverts. The Food Magazine, 9(1).
Food Commission (1992). A diet of junk food adverts – part 2. The Food Magazine, 18(2).
Hastings G et al. (2006). The Extent, Nature and Effects of Food Promotion to Children: A Review of
the Evidence. Technical Paper prepared for the World Health Organization. Geneva, World Health
Organization.
Hastings GB et al. (2003). Review of Research on the Effects of Food Promotion to Children. Report
to the Food Standards Agency. Glasgow, University of Strathclyde, Centre for Social Marketing.
Hawkes C (2006). Uneven dietary development: linking the policies and processes of globalization
with the nutrition transition, obesity and diet related chronic diseases. Globalization and Health, 2:4.
Institute of Medicine (2006). Food Marketing to Children and Youth: Threat or Opportunity? Institute
of Medicine of the National Academies, Washington DC.
Khan KS et al., eds. (2001). Undertaking Systematic Reviews of Research on Effectiveness. CRD’s
guidance for those carrying out or commissioning reviews. York, NHS Centre for Reviews and
Dissemination, University of York.
Lee M, Ulgado FM (1997). Consumer evaluations of fast-food services: A cross national comparison.
The Journal of Social Marketing, 11(1):39–52.
LODESTAR
(2002).
How
Cricket
Works.
A
LODESTAR
media
(http://www.excahnge4media.com/e4m/others/e4m_bt.asp, accessed 4 April 2006).
study
McDermott L et al. (2004). Desk Research to Examine the Effects of Food Marketing on Children.
Report to the World Health Organization. Glasgow, University of Strathclyde, Centre for Social
Marketing.
McNeal JU, Yeh C (1997). Development of consumer behaviour patterns among Chinese Children.
Journal of Consumer Marketing, 14(1):45–59.
49
Ofcom (2004). Childhood Obesity - Food Advertising in Context: Children’s food choices, parents’
understanding
and
influence,
and
the
role
of
food
promotion
(http://www.ofcom.org.uk/research/tv/reports/food_ads/report.pdf, accessed 9 April 2009).
Ofcom (2006). Television advertising of food and drink products to children
(http://www.ofcom.org.uk/consult/condocs/foodads/foodadsprint/section2.pdf, accessed 4 April 2006).
Vignali C (2001). McDonald’s: ‘think global, act local’ - the marketing mix. British Food Journal,
103(2):97–111.
Watson JL (2000a). Food as a lens: The past, present, and futures of family life in China. In: Jing J,
ed. Feeding China’s Little Emperors: Food, Children, and Social Change. Stanford, Stanford
University Press, 199–239.
Watson JL (2000b). China’s Big Mac attack. Foreign Affairs, May/June, 79(3):120–134.
Which? (2006). Food Fables. Exploding industry myths on responsible food marketing to kids.
Campaign Report. London, Which? (Consumers’ Association).
Wiener N (2004). How to engage with today’s kids. Advertising and Marketing to Children, 5(4):46–
52.
Witkowski TH, Ma Y, Zheng D (2003). Cross-cultural influences on brand identity impressions: KFC in
China and the United States. Asia Pacific Journal of Marketing and Logistics, 14(1/2):74–88.
50
ANNEX 1
Record of Searches
Academic Databases
Business Source Elite (15.11.08)
Searched: titles and abstracts
Limits: January 2006 – November 2008, Scholarly (Peer Reviewed) Journals
Search string
Results
(TI children and TI (food OR diet OR nutrition) and TI (marketing OR 23
advertising OR promotion)) OR (AB children and AB (food OR diet
OR nutrition) and AB (marketing OR advertising OR promotion))
CSA Illumina (includes Eric, Medline and Sociological Abstracts) (15.11.08)
Searched: keywords (Title, Abstract, Descriptors) in All databases
Limits: 2006-2008
Search string
Results
KW=children and KW=(food or nutrition or diet) and KW=(marketing 547
or advertising or promotion)
Cochrane Library Search (15.11.08)
Searched: title, abstract and keywords
Limits: 2006-2008
Search string
Results
(children):ti,ab,kw and (food):ti,ab,kw and (marketing or advertising or 72
promotion):ti,ab,kw
EBSCOHost (including International Bibliography of the Social Sciences and
PsycINFO) (15.11.08)
Searched: titles and abstracts
Limits: January 2006 – November 2008
Search string
Results
(TI children and TI (food OR diet OR nutrition) and TI (marketing OR 125
advertising OR promotion)) OR (AB children and AB (food OR diet
OR nutrition) and AB (marketing OR advertising OR promotion))
A-1
Emerald (15.11.08)
Searched: All fields (excluding fulltext), all content (not just subscribed content)
L: 2006-2008
Search string
Results
children AND (nutrition OR diet OR food) AND (marketing OR 20
advertising OR promotion)
Medline via PubMed interface (15.11.08)
Searched: title and abstract
Limits: added to PubMed in the last 30 days
Search string
Results
children[tiab] AND (nutrition[tiab] OR diet[tiab] OR food[tiab]) AND 4
(marketing[tiab] OR advertising[tiab] OR promotion[tiab])
Web of Knowledge (including Science Citation Index Expanded, Social Science
Citation Index, Arts & Humanities Citation Index) (15.11.08)
Searched: Topic (Title, Abstract, Author Keywords, Keywords Plus®)
Limits: 2006-2008
Search string
Results
Topic=(children AND (nutrition OR diet OR food) AND (marketing OR 323
advertising OR promotion))
Online Databases and Websites
Advertising Education Forum (23-24.03.09)
http://www.aeforum.org
Search term
2006
2007
2008
Results
136
99
69
A-2
Browsed database by “country” (country names as listed in the database). Reviewed
all the countries in the database classified as having a low-income, lower-middle
income, middle-income or upper-middle income economy according to World Bank
Classification (http://go.worldbank.org/K2CKM78CC0).
Countries
Argentina
Bolivia
Brazil
Chile
China
Cuba
Ecuador
India
Iran
Latvia
Lesotho
Lithuania
Malaysia
Mexico
Nepal
Pacific Islands
Pakistan
Peru
Philippines
Poland
Romania
Russia
South Africa
Sri Lanka
Taiwan
Turkey
Uruguay
Venezuela
ESRC Society Today (Economic and Social Research Council) (23.03.09)
http://www.esrcsocietytoday.ac.uk/ESRCInfoCentre/AdvancedSearchPage1.aspx
Searched: “ESRC Awards & Outputs”, sorted results by date.
Limits: reviewed the first 106 results (2006-present)
Search string
Results
(children AND (advertising OR food OR marketing)) OR (food AND 106
advertising)
WHOLIS (World Health Organization Library and Information Networks for
Knowledge Database) (23.03.09)
http://dosei.who.int/uhtbin/cgisirsi/4MJBx1LO6k/21350188/38/1/X/BLASTOFF
Limits: publication year 2006-2008
Search string
Results
(children AND (advertising OR food OR marketing)) OR (food AND 7
advertising)
A-3
LexisNexis Business and News (23.03.09)
Source: All English Language News
Limits: publication year 2006-2008
Search string
(ATL3(children) AND ATL3(food ) AND ATL3(advertising OR
marketing OR promotion)) AND Subject Type “Reports, Reviews &
Sections”
(ATL3(children) AND ATL3(food ) AND ATL3(advertising OR
marketing OR promotion)) AND Subject Type “Research Reports”
(ATL2(children) AND ATL2(food ) AND ATL2(advertising OR
marketing OR promotion)) AND Geography “Africa”
ATL2(children) AND ATL2(food ) AND ATL2(advertising OR
marketing OR promotion)) AND Geography “Asia”
ATL2(children) AND ATL2(food ) AND ATL2(advertising OR
marketing OR promotion)) AND Geography “Middle East”
ATL3 – At least 3 mentions; ATL2 – At least 2 mentions
Food and Agriculture Organization of the United Nations
FAO Corporate Document Repository (23.03.09)
http://www.fao.org/documents/
Browsed by “topic” and sorted by year of publication:
• Nutrition and consumer protection
• Trade and marketing
New Internationalist (23.03.09)
http://www.newint.org/issues/
Browsed the 2006-2008 issues
A-4
Results
69
65
12
36
4
ANNEX 2
Data Extraction Table: The Extent and
Nature of Food Promotion to Children
Shading indicates a new study published between 1st March 2006 and 15th November 2008
STUDY
Aktaş
2006
Arnas,
Alexander et al.,
1998
COUNTRY
Turkey
STUDY DESIGN
Content analysis
SAMPLE
5 weeks of TV programmes
and adverts broadcast 07:00
to 11:30 hours on Saturdays
and Sundays were watched.
Every week only one channel
was watched and
the data recorded on a semistructured observation form.
KEY RESULTS
The time devoted to children's programs was approximately 121 min and the
adverts during this period were approximately 35 min. A total of 344 of the 775
television adverts shown were related to food. Most of the food adverts were
about candy/chocolate, chips, milk and milk products such as cheese, yoghurt,
and breakfast cereals.
USA
Content analysis
75 advertisements yielded
from 24 archived shows from
the 1950s
Time trends analyses for the 1950s, 1970s, and 1990s are undertaken by citing
Barcus & Wolkin (1977) and Kunkel & Gantz (1992). The current analysis finds
that adverts are longer; > 1 min compared with around 30 s in earlier decades.
Advertised product categories in children’s programs are reported as: toys 8.3%,
cereals 23.3%, candy/snacks 21.7%, fast-food 0% and other products 46.7%.
Categories in all programs: toys 6.7%, cereals 20.0%, candy/snacks 17.3%,
fast-food 0% and other products 56.0%.
The study also attempts to look at promotional techniques and reports that
across children’s television from three decades, live action in advertising has
fallen (the article states that the 1950s were indisputably the decade of the adult
white male spokesperson) from 70.0% to 58.0% to 55.9%. Meanwhile animation
in advertising has increased only from 1.7% to 16.0% and 17.0%. The analysis
also notes a rise in the use of disclaimers from 8.3% to 41.0% and to 51.1% in
the 1990s.
A-5
The study concludes that by 1959, dominance of the “Big Four” product types in
children’s advertising had not occurred. The 1950s are described as an
important period for the introduction of brand building as this period marked the
origin of Kellogg’s Tony the Tiger character and the ‘Snap! Crackle! and Pop!’
slogan.
Alvy & Calvert,
2008
USA
Cross-sectional
content analysis
10 popular children’s websites
from
KidSay’s
market
research with 8-11 year olds
in Feb 2005
Seven of the 10 websites contained food marketing.
2
Instances of food marketing and their characteristics from the sites underwent χ
analysis. Of 15 possible food categories, the only marketing observed was for
candy (248 instances), sweetened breakfast cereals (42), quick-serve
restaurants (9), chips (3), dairy products (3), other (2) and sweet snacks (1).
All advergames and integrated marketing pages contained bold/colourful text
and dynamic images. 98.3% of adverts used bold/colourful text and 90.2% of
adverts used dynamic images. Animation was used heavily in all marketing in all
but product placements.
One in 7 of the marketing instances used a branded character.
Repetition was prevalent: 112 of 308 food marketing instances were unique.
Only 32 products and 18 brands were marketed overall.
One website was a food product site and thus contained a significantly greater
amount of food marketing than the other children’s sites.
Atkin, 1975a;
Atkin & Heald,
1977
USA
Longitudinal content
analysis
470 adverts taken from 4
hours of children’s Saturday
morning television
The content analysis provides a breakdown of advertising by product for both
1972 and 1973:
• 1972 – toys comprised 50% of advertising, “especially sugared” cereals
comprised 27%, and candies/drinks/sweets/snacks/deserts/fast-food
comprised 21% of advertising (total for food advertising, 48%).
• 1973 – toys comprised 66% of advertising, “especially sugared” cereals
comprised 17% of advertising, and
candies/drinks/sweets/snacks/deserts/fast-food comprised 15% of advertising
(total for food advertising, 32%).
The study shows that the reduction in adverts during the heavy toy selling
season was accounted for by reducing food advertising. Nevertheless, after
toys, food was the top product advertised. Cereals (especially sugared cereals)
were the most advertised food product, followed by candies, drinks, snacks
sweets, deserts and fast-foods.
Creative strategies were also analysed by comparing the role models in each
type of advert. It was noted that mothers tended to appear in food adverts, and
that adults were only portrayed as heroes and villains in food adverts (toy
adverts tended to feature children alone). Theme appeals were also assessed
showing that 98% of toy adverts took a serious appeal, and 92% of food adverts
took a humorous appeal. Much of these data are not adequately broken down by
A-6
food or food product, and reporting is somewhat subjective, e.g. “food adverts
almost exclusively based on the fun claim, while toy adverts frequently
emphasized feelings of power and being grown up”.
Only food adverts made use of animation and these were more likely to adopt a
humorous (fun as opposed to serious) tone. Of food adverts, 62% included at
least some animation, compared with only 1% of toy adverts. Of food adverts,
94% adopted a fun tone, compared with 43% of toy adverts.
The study also examines nutritional appeals and mentions sweetness (21%) and
nutritional claim (47%). Only 2% of adverts used “tell mom to get this cereal”, i.e.
explicit ‘pester-power’ strategies in 1972. Both food and toy commercials are
said to rely on celebrity endorsement/testimonials or on more general customer
satisfaction to a “limited extent”. In contrast, 24% of food adverts used
premiums.
Bang & Reece,
2003
USA
Content analysis
813 adverts in children’s
television programming
Caucasians continue to be the predominant models in terms of the types of roles
that they play in food advertising and how much they feature. The representation
of minority groups has improved, though they are still more likely to play minor
roles in advertising.
Of all adverts featuring black models, 61.1% showed them in food adverts
(compared with 46% for Caucasian models).
Barcus,
1971b
1971a;
USA
Content analysis
311 adverts from 4 channels
of
Saturday
morning
television
The content analysis provides a breakdown by product category - toys (22%),
candies/sweets (20%), cereals (24%) and other food snacks (22%). In terms of
creative strategies, all toy adverts were non-animated, although 66% of food
adverts were animated or mixed. Of toy adverts, 92% showed the product in
use, while only 72% of food adverts showed the product in use. Food adverts
usually followed a dramatic skit (59%) while toy adverts used off-stage voice
(73%).
Cereal adverts were more likely to make nutritional claims, and 19 out of 36
adverts mentioned sweetness appeal. Only four in 21 adverts for
sweets/candies/soft drinks (and five in 32 adverts for other foods & snacks)
made nutritional claims, with a baseball player making an energy claim
endorsement for a product. Of 21 adverts for sweets/candies/soft drinks, 17
showed the child consuming the product.
Disclaimers were rarer with foods adverts (verbal, n = 8, 6 for cereals, compared
with 10 for toys and visual, n = 3, for all cereal adverts, compared with 5 for
toys). Also mentions some inappropriate marketing, though this is not food
focused.
A-7
Barcus, 1975a;
1975b,
Barcus
&
Wolkin, 1977
USA
Partly
longitudinal
content analysis
The sample of adverts for the
content
analysis
of
commercial
television
at
weekends was taken from
television during April 1975,
which yielded 403 adverts
(137 versions of 119 items for
65 sellers (98% of all adverts
of 30 s length))
Children’s commercial television on the weekends:
Content analysis of top advertised products showed that cereals and
candies/sweets both comprised 24.8% of advertising, then toys at 18.1%. Other
food groups advertised being eating places/meals (10.4%), snacks (4.0%) and
other food (4.0%). Total of food advertising = 68.0%. The study states a ratio of
3:1 for advertised cereals of the sugared variety (n = 76:23) and identifies few
adverts for milk/dairy products (1.5%), bread (0.5%) or fruit/juices (1.2%). No
adverts were for fresh meat or vegetables.
A breakdown of animation strategies, mixed animations strategies and nonanimation strategies by product category is provided (in percentages).
Respectively, these are: cereals, 18, 62, 20; candies/sweets, 22, 29, 49; eating
places/meals, 0, 0, 100; snacks, 25, 6, 69; other foods, 25, 6, 69. The analysis
showed that, with the notable exception of restaurants, food adverts often
included at least some animation whereas toy adverts did not. Toy adverts also
tended to have an off-screen announcer format.
A breakdown of off-screen, on-screen, musical and drama formats by product
category is provided (in percentages). Respectively, these are: cereals, 3, 17,
29, 51; candies/sweets, 21, 10, 32, 37; eating places/meals, 10, 29, 52, 9;
snacks, 0, 38, 0, 62; other foods, 31, 12, 19, 38. The analysis found that a
dramatic skit was the most common method used in food adverts, with the
exception again being restaurants where music (e.g. McDonalds jingle) was
more common. A breakdown of the status of ‘product display’ in advertising in
terms of “shown”, “in use” and “name only” (remainder = unclassifiable others)
was also provided by product category (in percentages). Respectively, these
are as follows: cereals, 46, 50, 0; candies/sweets, 18, 76, 6; eating
places/meals, 5, 50, 24; snacks, 6, 94, 0; other foods, 44, 56, 0. The analysis
found that fast-food restaurant advertising was less reliant on showing the
product, (i.e. they tended to rely more on branding) and that cereal advertising
tended to feature the box rather than the product in use.
The study also examines ‘who speaks’ for the product. Of toy adverts, 100%
were reported as featuring adults (with 75% of these being male). A breakdown
of the percentages of adults and children featuring in advertising by product
category is also provided. Respectively, these are as follows: cereals, 55, 19;
candies/sweets, 66, 6; eating places/meals, 98, 2; snacks, 6, 0; other foods, 44,
50. No product advertising was majority female. The eating place category again
seems to be the exception in its use of adults (like toys). The study observed a
small use of celebrity endorsement (7% of all adverts), the only food adverts
using personalities reported as being cereals (17%) and candies (7%).
A breakdown of the percentages of audio, visual and both audio and visual
A-8
disclaimers featuring in advertising by product category is also provided.
Respectively, these are as follows: cereals, 24, 15, 13; candies/sweets, 12, 10,
10; eating places/meals, 5, 10, 5; snacks, 16, 0, 0; other foods, 16, 0, 0. In
comparison, toys were: 55, 14 and 0 respectively.
The sample of adverts for the
content
analysis
of
commercial television after
school was taken from
television during May 1975,
which yielded 487 adverts
(1262 versions of 218 items,
80% of all adverts were of 30
s in length)
A breakdown of the percentages of premiums and contests featuring in
advertising by product category is also provided. Respectively, these are as
follows: cereals, 47, 0; candies/sweets, 10, 6; eating places/meals, 21, 5;
snacks, 16, 0; other foods, 16, 0. In comparison toys were 0 and 0 respectively.
A breakdown of theme appeals in terms of appearance, amount, convenience,
taste, texture, fun, health, status/superiority, action/adventure, comparative,
uniqueness/newness and quality is provided by product category (in
percentages). Respectively, these are as follows: cereals, 3, 3, 0, 36, 12, 2, 34,
2, 7, 1, 0, 0; candies/sweets, 3, 6, 5, 30, 19, 15, 4, 3, 0, 0, 15, 0; eating
places/meals, 2, 0, 0, 26, 22, 5, 0, 38, 0, 5, 2, 0; snacks, 0, 0, 17, 46, 17, 0, 0, 0,
0, 20, 0, 0, 0; other foods, 0, 3, 12, 37, 18, 0, 21, 0, 0, 9, 0 and 0. In comparison
toys were 19, 5, 9, 4, 0, 9, 0, 1, 44, 0, 1, 3 and 15 respectively. In this instance,
food advertising tended to focus on taste (except for cereals advertising where
nutrition was top). Those foods, e.g. candy, which lacked this claim opted for fun
or texture as secondary selling points. However, peer status, popularity or
general superiority were important with fast-food outlets
A range of examples of misleading adverts is given, but the study states that
more research is required. The study provides a breakdown of nutritional
adverts. The top advertised product is given as sugared cereals (27.8%),
followed by candy bars (17.2%), eating places (14.6%), non-sugared cereals
(8.8%), cakes & cookies (8.0%) and fruit drinks (7.3%). The study refers to
public service announcements as non-commercial announcements (n = 92) of
which nine (9.8%) were for nutritional messages.
Children’s commercial television after school:
The content analysis provides a breakdown of advertising by product category:
the top advertised products were toys (18.5%), then cereals (17.4%) and
candies/sweets (14.9%); other food groups being eating places/meals (11.3%),
snacks (0.4%) and other food (1.8%). Total of food advertising = 46.0%. The
ratio of sugared cereals to unsugared cereals is again reported as 3:1 (n =
65:20). In terms of creative strategies the study looks again at ‘who speaks’ for
the product; 100% of toy adverts featured adults (92% male). The percentages
of adults and children featuring in food adverts for different products are also
provided. Respectively, these are reported as: cereals, 37, 14; candies/sweets,
66, 14; eating places/meals, 95, 0; snacks/other foods, 91 and 9. Again no
product was majority female (although cereals featured the most females at only
17%).
A-9
This part of the study looks at who appears in the advert and reports a different
pattern in which toy adverts featured the most children. The percentages of
adults to children featuring in toy adverts for different products are also provided.
Respectively, these are: cereals, 13, 55; candies/sweets, 32, 24; eating
places/meals, 38, 50; snacks, 60, 0; other foods, 44 and 56. In comparison, toys
were 12 and 88, respectively. Again no product was majority female (other foods
most at 44%). The analysis suggests that for fast-food advertising, adults
announce, but children appear.
The sample of adverts used
to assess seasonal variations
in television advertising to
children was taken from
television during November
1975. The study compares
the 1200 min recorded in April
with 960 min recorded in
November, seen as run-up to
Christmas
The study also breaks down products by estimated audience of children viewing
programmes, into 0–29, 30–49, 50–69 and ≥ 70%. Respectfully, these figures
for food products are: cereals, 5, 14, 38, 43; candies/sweets, 9, 26, 40, 25;
eating places/meals, 11, 14, 29, 46; snacks/other foods, 18, 46, 9 and 27. Toys
were 3, 4, 37 and 56 respectively. This shows that foods were targeted at
children’s programmes, beaten only by toys. In comparison the percentages of
household products (33, 39, 11, 17) and personal care products (26, 42, 21, 11)
were more evenly spread and slanted in the opposite direction.
Again only a small use of celebrity endorsement is observed (5% of all adverts),
foods using personalities being cereals 4%, candies 0%, eating places 4% and
snacks/other foods 27% (again only cereal using endorsement 1%). No food
adverts were reported as using tie-ins.
A breakdown of the use of audio, visual and both audio and visual disclaimers
for different products is also provided (in percentages). Respectively, these are:
cereals, 0, 16, 16; candies/sweets, 1, 0, 0: eating places/meals, 22, 11, 6;
snacks/other foods, 0, 0 and 0. In comparison toys were 51, 1 and 1,
respectively. A breakdown of the use of premiums in food advertising across
different food product categories is also provided: cereals, 33%; candies/sweets,
1%; eating places/meals, 31%; snacks/other foods, 9%. No food contests. In
comparison toys 0% premiums, but 1% contests.
In terms of price information, only advertising for eating places used this 18.2%
(toys were 22%). Advertising for eating places were the least likely to provide
product information (especially when compared to advertising for household
products and personal care products).
The study also looks at misleading advertising and provides percentages broken
down by nutrition, “sweetness”, product is a snack, product is a meal, taste
(other than “sweetness”), and product is natural/artificial. The respective
percentages for the advertising of different food products are as follows: cereals,
41, 22, 0, 94, 74, 4; candies/sweets, 11, 0, 14, 8, 49, 3; snacks/other foods, 18,
A-10
0, 9, 9, 36, 9.
The study also provides a more detailed breakdown of nutritional adverts. The
top advertised products are given as; sugared cereals (29.0%), candy bars
(22.8%), eating places (11.2%), non-sugared cereals (8.9%), cakes & cookies
(7.2%) carbonated beverages (5.4%) and fruit drinks (4.0%). The report states
that 61% of all advertised foods are sugared.
A breakdown of theme appeals in food advertising in terms of appearance,
amount, convenience, taste, texture, fun, health, status/superiority,
action/adventure, comparative, uniqueness/newness and quality is also provided
by product category. The respective percentages are: cereals, 7, 3, 0, 36, 10,
1, 34, 1, 8, 0, 0, 0; candies/sweets, 7, 16, 2, 35, 11, 10, 7, 3, 0, 1, 6, 0; eating
places/meals, 0, 11, 20, 14, 9, 3, 0, 9, 0, 0, 4, 2; snacks, 0, 0, 50, 50, 0, 0, 0, 0,
0, 0, 0, 0; other foods, 8, 0, 8, 25, 0, 8, 17, 0, 0, 0, 8, 12. In comparison toys
were 17, 13, 6, 2, 2, 16, 0, 15, 23, 0, 5 and 1 respectively. Again advertising of
cereals was the biggest user of nutrition appeals. Only 52 public service
announcements (non-commercial advertisements) were identified, and only one
of these was for a nutritional cause.
Comparisons between food advertising to children at weekends and food
advertising after school showed that the advertising was similar (although more
non-child adverts were shown after school). Of nutritional adverts 4% were
shown at weekends, and 2% after school.
Seasonal variations in television advertising to children:
Advertising time was found to rise from 12.8–15.0%, and public service
announcements fell from 3.1–0.9%. The content analysis shows a rise in toy
advertising from 17.3–47.5%. Advertising of food products is shown to fall:
cereals, 23.5–19.7%; candies/sweets, 27.2–19.7%; snacks, 5.2– 0.3%; eating
places, 6.2–5.3%; other foods 3.6–1.8%. The examination of food advertising
alone revealed little change, with 142 of 200 adverts in April and 100 of 133 at
Christmas-time being for highly sugared products.
Specifically, variations in top advertised products between April and November
were sugared cereals, 29.5–33.1%; candy bars, 22.5–22.6%; eating places,
8.5–11.3%; cakes/cookies, 8.0–12.0%; fruit drinks, 7.0–1.5%; and unsugared
cereals, 6.0–9.0%. A breakdown of changes in the use of audio, visual and both
audio and visual disclaimers in food advertising across different product
categories is also provided. The respective figures show a rise in their use in
cereal advertising: 26–50%, 18–45% and 18–34% (in line with toys, 49–53%, 0–
18% and 0–13%).
A-11
Barcus, 1981
USA
Review article
‘New’ data consists of 33
hours of children’s television
The review states that the average child is exposed to 15 adverts, five promos
and two public service announcements per hour. The content analysis found
more food advertised as a percentage of total advertising on NAB-compliant
channels compared with independents. The reported breakdown by product are
as follows:
• NAB: cereals, 34%; candies/sweets, 29%; eating places, 15%; toys, 12%;
other food, 1% (food total = 82%).
• Independents: toys, 49%; cereals, 13%; candies/snacks, 13%; eating
places, 9%; other food < 0.5% (food total = 35%).
• Overall (133 adverts analysed): cereals, 24%; candies, 21%; and 12% for
fast-food such as McDonald's and Burger King (food total still top at 60%, n
= 133).
The review states that 70% of foodstuff adverts are for highly-sugared products,
and that less than 1% are for healthy products such as “meats, bread, dairy
products, fruits or vegetables”.
Creative strategies are also examined and three types reported:
i) “Product presentation” – Animation/live mix reported on 80% of cereal
adverts (e.g. Kellogg’s Tony the Tiger) compared with an off-stage
announcer in toy advertising (usually adult white male). Only 4% of food
adverts used live announcers and only 5% used testimonials. No explicit
evidence of overtly urging "pester power". Premium offers were reported
in 25% of cereal adverts.
ii) “Attention-getting devices” – More than half of all food adverts and 90%
of cereal adverts employed fantasy techniques (e.g. magical kingdoms
inside the box).
iii) “Qualifiers, disclaimers and disclosures” – Since 1971, many more
qualifiers were used. However, these can be misleading – inappropriate
marketing – e.g. cereals with added vitamins “part of a balanced
breakfast” could be construed as a “necessary part” of a balanced
breakfast.
The review reports little use of hard product information (e.g. price, quality or
ingredients): tend to be advertised more on taste or texture (added vitamins
aside). In “many cases” content could only be discerned from product name
(e.g. Corn Flakes). However, sweet sugary nature often mentioned.
Four types of verbal appeals are also listed:
i) “Assertions” – e.g. “will save money” is “fun” or “convenient”; only about
30% (of all children’s adverts) used this. Fun usually used and sometimes
convenience.
ii) “Attributed Qualities” – e.g. “tastes great” or “country fresh”. “Almost nine
A-12
out of ten” food adverts used qualities which the observer may not agree
with (e.g. taste in 80% and novelty in 10%).
iii) “Product Properties” – e.g. size, shape, colour or texture; 70% of food
adverts had some of this. The major property in food adverts was texture
(e.g. crunchy/chewy) which rose to 60% with cereal adverts.
iv) “Product Composition” – e.g. ingredients (what the product was made of);
this was found in 90% of cereal adverts. About 40% of adverts had this in
their product name alone, with others including vitamins, honey, sugars,
flavours and chocolate.
The author claims that visual appeals were of hedonism (e.g. adventure and fun)
rather than education, work or relationships (only two of 133 food adverts took
place in a work setting).
Barnabè et al.,
2008
Italy
Review
Existing data and expertise,
particularly relating to Europe.
EXTENT
Children watch TV for an average of 3 hours/day. Children in low or middle
income working families and where both parents are out of the home most of the
day watch the highest exposure of 6 hours/day.
Food is the main sector in Europe in terms of advertising expenditure. Available
data for France, the United Kingdom, Italy, Germany and Spain show that, in
children airtime (1400-1800 hours), one food commercial is broadcast every 10
min (every 5 in Italy). This translates to 33,000 commercials per year.
About 60% of food advertising is programmed to air between 1600 and 2100
hours and about 40% of TV adverts for big-6 category products (soft drinks,
confectionary, snacks, fast food, convenience food, cereals) is in children’s
airtime. Thus children watch an average of 29 commercials per day and in
children’s airtime (1400-1800 hours), 60% are for Big-6 food category products.
Between 1400 and 2100 hours, 40% of adverts are for big-6 products.
The authors could not identify any equivalent market research about children’s
exposure to other advertising media (outdoor, print, cinema, radio and internet)
although in Europe the media mix for the whole food sector is 85% TV, 10%
press and the rest outdoor and radio, with a growing internet budget.
The prevailing food and beverage marketing messages to children and youth
have focused on products that are high in total calories, sugar, salt, and fat and
low in nutrients. The advertised diet contrasts sharply with the one
recommended by public health organizations; and there is little promotional
support for healthy dietary behaviour.
NATURE
A-13
In the food market in general, TV advert creative approaches are usually based
on the genuineness and naturalness of ingredients, the nutritional and
nourishing values, and the consumer’s similarity to the people/situation in the
advertising. The creative treatments are calibrated towards the ideal consumer
and developed through cartoon (typical children’s world and interaction between
the hero and children); lifestyle (product is the means for having a good
time/finding friends/being trendy); spending time with the family.
Other types of food marketing to children include promotions, sponsorship, point
of sale, gifts in pack, prizes, co-marketing (amusement parks/movies). Recently
an approach based on charity or co-marketing with a non-profit organization has
been widely used. Other new methods include on-screen advertising on buses
and trains, advertising screens in stores, hyper-tag technology using Bluetooth
ports of mobile phones and PDAs, interactive digital billboards, chatbots
(marketing AI software programs in chatrooms) and blogs.
Batada
&
Wootan, 2007
Batada
2008
et
al.,
USA
USA
Cross-sectional
content analysis
Cross-sectional
content analysis
Nickelodeon
(the
mostwatched children’s television
station)
media:
NICKI
television channel (28 hours –
23-24
September
2005),
Nickelodeon Magazine (4
monthly issues September to
December 2005), all products
with Nickelodeon characters
on the pack in large grocery
store in Washington DC (3day period in September
2005).
Nutritional quality was assessed using set of standards created by a panel of
nutritionists and health experts based on fat, sugar, salt, nutrient content, and
meal portion sizes.
27.5 hours Saturday morning
children’s
television
programming from the major
broadcast and cable networks
on the same day in May 2005
4.08 hours of advertising were broadcast including 1.99 hours of food adverts.
Of 572 adverts, 281 (49%) were for food. The most heavily marketed foods and
beverages, by category, were ready-to-eat breakfast cereal and cereal bars
(27% of food adverts), restaurants (19%), snack foods (18%), and candy (14%).
Of 168 television food ads, 148 (88%) were for foods of poor nutritional quality.
Of 21 magazine food ads, 16 (76%) were for foods of poor nutritional quality.
Fifteen grocery store products were identified with Nickelodeon characters on
the packaging; nine (60%) were foods of poor nutritional quality.
Of the 48 possible children’s meal combinations at restaurants with promotional
offers tied to Nickelodeon programs, 45 (94%) were of poor nutritional quality.
Of the 281 food adverts, 91% of the foods were high in fat, added sugars, or
sodium, or low in nutrients.
The most commonly used marketing techniques were movie, cartoon, animated,
or costumed characters (used in 74% of food adverts), giveaways (26%), use of
websites/e-mail (15%), and animation (15%).
86% of adverts had an emotional appeal, such as fun or being hip or cool.
A-14
Berry
&
McMullen, 2008
Canada
Cross-sectional
content analysis
Sample of 15 supermarket
cereal aisles in Toronto –
images taken 48 inches from
the ground and 24 inches
back from the shelves.
Breakfast cereal products with higher-than-average levels of sugar, refined
grains and trans-fats were more likely to feature child-oriented marketing in the
form of spokes-characters, themed cereal shapes/colours and child incentives
on cereal boxes.
Cereals shelved within reach of children aged 4-8 years had less sugar per
serving and were less likely to contain trans-fats than less reachable products.
Buijzen
Valkenburg,
2002
&
USA
Content analysis
601 adverts from 216 hours of
recorded
television
from
children’s programming
The content analysis found that the top five products in children’s adverts were
toys & games adverts (58.3%), followed by candies and snacks (12.2%), music
& video (12.2%), non-alcoholic drinks (5.8%) and foods (5.0%). In comparison,
soft drinks were in the top five teenage products advertised (11.9%) and foods in
the top five general audience products advertised (6.5%).
The creative strategies of the advertisements were assessed using an appeal
coding system. Appeals featured to the following extent in advertising to
children: play (57.6%), action-adventure (38.8%), fun (30.2%), courage (7.2%),
affection for animals (12.2%), collecting (6.5%), nurturing (8.6%), creativity
(6.5%), affection for children (9.4%), competition (8.6%), family ties (5.0%) and
capability (2.2%). Appeals featured to the following extent in advertising to
teenagers: being modern (18.3%), being ‘cool’ (18.3%), fun (20.6%), seizing
opportunities (20.6%), sexuality (13.5%), individuality (9.5%), personal freedom
(11.1%), having the best (23.0%), belonging to a group (9.5%), energy (5.6%)
and enjoyment (6.3%). Appeals featured to the following extent in advertising to
the general audience: convenience (17.9%), financial security (11.0%), health
(11.6%), sexuality (10.7%), love (6.5%), individual (5.7%), physical
attractiveness (14.6%), personal freedom (7.4%), affection for children (7.7%),
family ties (7.7%), having natural organic food & clothing (3.9%), self-esteem
(3.3%), enjoyment (8.0%) and career (2.4%).
The top five appeals used in the advertising of candy & snacks were pleasant
taste (88.6%), humour (85.7%), newness (31.4%), fun (17.1%) and actionadventure (14.3%). The top five appeals for the advertising of non-alcoholic
drinks were humour (83.7%), pleasant taste (34.9%), energy (34.9%), newness
(20.9%) and fun/quality (18.9%). The top five appeals for adverts for foods were
humour (82.5%), taste (82.5%), quality (27.5%), newness (22.5%) and health
(22.5%).
The study concludes it is through adverts for toys or candy that children are
“confronted with specific child-related appeals, such as play, fun, actionadventure and humour”.
A-15
ByrdBredbenner
&
Grasso, 1999a;
1999b; 2000a;
2000b; 2000c
USA
Content analysis
700 adverts taken from 17.5
hours of prime-time television
Recommended Diet
The USDA Food Pyramid is not fats, oils & sweets 41%, milk, cheese and
yoghurt 8%, meat, poultry, fish, dry beans & eggs group 0%, vegetable group
6%, fruit group 6% and bread, cereal, rice & pasta group 41%.
The study simply states that “fruits, vegetables (except French fries) and dairy
products were rarely advertised” and “protein rich foods and grain products were
well represented in prime-time advertising mainly because of frequent
advertisements for fast-food sandwiches”. A brief mention is made of foods in
the fats and sweets group in the USDA food guide pyramid being advertised
frequently. References (n = 755) to ‘low nutrient density foods’ coupled with
foods in the ‘foods high in the fats, sweets and alcohol’ exceeded the
references (n = 667) to high and moderate nutrient density foods combined. The
study concludes that the prime-time diet “can be described as calorie-laden,
fatty, salty, sweet and low in fibre”.
Public service announcements
The analysis of public service announcements was found to replicate previous
and other research as none were observed for diet. The study observes a mixed
message in that 89% of actors consuming foods in food adverts were slim and
healthy yet 54% of food consumed was rated low nutrient density. Only one PSA
with NRI was recorded; in this a child gives an anti-drug message whilst eating
French fries.
In comparing time trends with 1992, the study observes a rise in the amount of
advertising from 19–24% of television time. It also reports a rise in adverts with
health information from 246 to 298 (or 14 to 17/hour) (N.B. no mention of
significance). Content analysis of both 1992 and 1998, reported respectively 105
adverts (28%) and 108 adverts (23%) for food, of which 100% had health
information. This is the largest category in each. In both years only five healthorientated PSAs were recorded. In 1992 all five were about drugs and AIDS, and
in 1998 one was about domestic violence, the rest about drugs. The study also
monitored non-adverts, non-PSAs, and non-programme promotions, and states
that 30% of non-programme time is for these, with over 25% of these promotions
containing health information.
The proportion of advert time is reported as similar between 1992 and the
current study, at 79% (156 minutes) and 77% (199 minutes) respectively. This
was made up of 337 adverts and 467 adverts of which 204 (54%) and 231 (49%)
had health information. The study also claims that the amount of misleading
information in food advertising declined from 75% to just over 50%.
In comparing data from 1971, 1977 and 1990 with the current study, it is
A-16
reported that the percentage of food adverts is decreasing (from 31 to 35, to 20
to 15% respectively), but set against rise in total adverts (11 to 40 per hour), the
amount of exposure remains the same. Changes over time in products
advertised are significant (P < 0.001); dominated by food categories
“restaurants”, “low-nutrient beverages”, “protein rich foods” and “cereals”. The
study observes the main trend towards restaurants (fast-food, hamburgers, friedchicken and pizza) from “virtually non-existent” to being the largest category.
The study also looks are creative strategies and states that almost half of food
and drink adverts made misleading or inaccurate claims. A total of 42 of the 108
food adverts made at least one nutritional claim, compared with 85 with taste
claims (and 97 with any consumer-related claim) In 40 adverts (2%) only a
restaurant name was given, not food product, providing evidence of brand
building. The study concludes that the prime-time diet is similar to the unhealthy
diet in the American population, i.e. high in sugar, sodium and fat but low in
fruits, vegetables and whole grains.
ByrdBredbenner,
2002
USA
Review article
Adverts
sampled
from
recordings in both 1993 and
1999 which yielded 11.5 and
9.5 hours respectively of
Saturday morning television
This article is the most up-to-date in a series of content analyses from the
author. The total of food adverts (as a % of all adverts) was 70% in 1971, 69% in
1993 and 78% in 1999. Product trends are also provided across the four data
points (1971, 1975, 1993 and 1999) (N.B. figures are in %):
• Fast-food restaurants’ advertising has increased from 8, 11, 23 to 28%.
• Fats and sweets advertising has fluctuated from 49, 41, 36 to 53%.
• Advertising for breads and cereals has fluctuated from 37, 45, 40 to16%.
The data for sweetened cereal advertising is different (27, 33, 24, 14%): in
other words advertising for cereals is decreasing, but sweetened cereals
much less so.
• Advertising for dairy products has fluctuated from 3, 2, 1 to 3%.
• Advertising for frozen dinners has decreased to zero (2, 0, 0 to 0%), as has
advertising for vegetables (1, 0, 0 to 0%).
• Advertising for fruit has remained at 0% throughout as has advertising for
high protein foods.
In terms of creative strategies, the review addresses inappropriate marketing –
and also looks at the characters who eat in adverts. In 1993, the proportion of
thin/average-sized characters was 81%; in 1999 this had risen to 96%.
Premiums and prizes were found to be (35% in 1993, 29% in 1999) similar to
1970s levels, however these had switched from breakfast cereals to fast-food
advertising (61% and 79%). The review also states that in 1993 only 48% of
advertising focused on the prize, whereas in 1999 this had risen to 87%.
A-17
Carter, 2006
Australia
Systematic literature
review
Research,
reviews
and
commentaries
identified
through Boolean searches of
PubMed, ScienceDirect and
Google;
searches
of
bibliographies of relevant
papers – with preference
given to Australian data.
Australia children are exposed to more television food advertising than the US,
UK, New Zealand, or 11 other western European countries. The average
estimate suggests 9-12 food adverts per hour on Australian children’s TV, with
74-99% of these being for energy-dense foods (e.g. fast foods, soft drinks,
lollies, ice cream, chocolates, and snack foods). Equates to the average
Australian child viewing 6,074 energy-dense food adverts per year, or 17 per
day.
Chapman,
Nicholas
&
Supramaniam,
2006
Australia
Cross-sectional
content analysis
672 hours of TV programmes
and adverts taped from all 3
commercial stations available
on
free-to-air
Australian
television between 0700 and
2100 hours on 2 weekdays
(Tue 7-Wed 8 June) and 2
weekend days (Sat 4-Sun 5
June) in 2005 in Sydney,
Brisbane, Tamworth (NSW)
and Ballarat (V).
645 hours were available for analysis. A total of 10,593 adverts were analyzed,
of which 3,287 (31%) were food adverts. 81% of the food adverts were for
unhealthy/non-core foods, which equated to 25% of all adverts screened
between 0700 and 2100 hours.
A similar number of health/core food adverts were screened in the metropolitan
2
and regional areas. A significantly higher proportion (χ = 29.7, P<0.0001 with 1
degree of freedom) of total and unhealthy/non-core food adverts were screened
in the metropolitan areas. An average of 4.13 unhealthy/non-core food
adverts/hour, more than four times the number of healthy/core food adverts
(0.96 adverts/hour)
Weekdays between 1800 and 2100 hours and Saturday between 0900 and 1100
hours were the most concentrated periods for advertising unhealthy/non-core
food (at over 5 adverts/hour and over 6 adverts/hour respectively). The early
time slot on Saturday was the most concentrated period for advertising
unhealthy/non-core food and significantly different from the weekdays and
2
Sunday levels at the same time (χ =6.89, P=0.03 2 df). Overall, the level of food
advertising shown on weekdays compared with weekends was identical.
Chapman et al.,
2006
Australia
Cross-sectional
content analysis
Food products from 7 food
categories using promotional
tactics (giveaway/competition,
cartoon
character,
sports/movie
celebrity,
character promotion) or not,
from 9 supermarkets (3 of
each chain) around Sydney in
socio-economically
similar,
family-oriented
shopping
areas. Data collected 11-18
August 2005.
A-18
All the surveyed supermarkets carried food products promoted using the
techniques surveyed. Confectionary had the greatest variety of products with
145-190 available. Confectionary had the highest proportion of promoted
products (35% average), then dairy snacks (31%), snack foods (30%).
Promotional methods were used less for breakfast cereals, ice creams (both
15%) and chips/savoury snacks (9%). These differences in proportions were
2
statistically significantly different (χ =29.8, P<.001 6 df).
100% of promotional activity within the confectionery, sweet biscuit,
chips/savoury, dairy snacks and ice cream categories directed at children. The
total number of promotions amounted to 231, of which 82% (n=189) were used
to market unhealthy foods and 18% (n=42) to promote healthy foods. Snack
food had the highest level of multiple food promotion usage on a single product.
Breakfast cereals, although they seldom used food promotions, used multiple
ones when they did. Most promotions were run singularly on confectionary
products. 48% promoted foods used cartoon characters unique to the food pack
(i.e. not from TV or movies) however this was not uniform across food types
2
(χ =30.5, P<.0001 6 df). 75% confectionary products used cartoon characters
and cereal companies tended to use well-known characters. 27% promoted
foods used TV of movie characters – predominantly cartoon-based with the
exceptions of The Wiggles and Star Wars. 100% promoted products in dairy
snacks, ice cream, sweet biscuits and chips/savoury snacks used cartoon or
TV/movie characters. 13% of promoted foods used giveaways. Snack foods
used this method the most when promoting a product (35%). 12% of promoted
foods used competitions. 50% of promoted cereals and 35% of promoted ice
creams used competitions. 5% of promoted cereals (one product) used sports
stars.
Chester
Montgomery,
2007; 2008
&
USA
Exploratory review
Examines US food and
beverage companies’ public
documents
and
press
statements, supplemented by
analyses of websites and
other online content, to
identify
the
major
contemporary strategies used
by food marketers to promote
brands to children and
adolescents.
A-19
The new marketing practices by food and beverage companies include:
Mobile (cell) phones enable companies to directly target users based on
previous buying history, location, and other profiling data.
Database marketing has become a core strategy, on the Internet, cell phones,
video games, and other new platforms. Profile compiled from demographic data,
purchasing behaviour, responses to advertising messages, and the extent and
nature of social networks.
Peer-to-peer/buzz/viral/word-of-mouth marketing used to target key influential
young people, ‘brand sirens’/‘brand advocates’, to promote products to peers
through instant messaging, social networking sites, and blogs. One campaign
across different media. Often use viral videos, with the sponsoring company
disguised.
The three major instant messaging formats (AOL, Yahoo! and MSN) offer a
variety of strategies, including ‘roadblocks’ and ‘takeover ads’ that flood a site’s
homepage with interactive commercials, as well as branded ‘bots’ and buddy
icons.
Commercializing online communities/social networking sites.
Food and beverage companies have created their own online branded
entertainment sites, designed to encourage young consumers to engage
playfully with products over long periods of time, many offer ‘free’ content,
games, merchandise, and endless replays of television commercials.
In-game advertising, or ‘game-vertising’, combines product placement,
behavioural targeting, and viral marketing to forge ongoing relationships
between brands and individual gamers.
Online immersive 3D environments or virtual worlds are complex, multilayered
enterprises that combine instant messaging, interactive gaming, and social
networking into increasingly elaborate settings, in which individuals create their
own online identities through avatars. Food companies brand and marketing
messages can be spread by avatars and displayed in the virtual world.
Chestnutt
Ashraf, 2002
&
United
Kingdom
Content analysis
237 hours of children’s
television and 42 hours of
prime-time television yielded
2345
and
891
adverts
respectively
The content analysis found significantly more food advertising during children
programmes (62.5%) than on prime-time television (18.4%). Moreover, of
advertising time devoted to food adverts during children’s programming, 73.4%
were for products deemed detrimental to oral health, compared with only 18.6%
similarly categorized during prime-time. The content analysis also provides a
breakdown in terms of products detrimental to oral health that are reported and
notes that the top foodstuffs include; confectionary (46.6%), sugared cereals
(24.1%), sugared-dairy products (16.0%), sugared soft-drinks (10.9%), diet
(1.3%) and low-sugar (1.2%).
The study also points out that the two Saturday and Sunday morning children’s
television shows broadcast in the United Kingdom are sponsored by a
confectionary and a sugared dairy product (tie-ins).
Childs & Maher,
2003
USA
Content analysis
215 advertisements
Nearly half (46%) of the advertising tapes was for food or candy.
A sample of food advertising to children exhibits greater gender preference in
presentation than a comparison sample of non-food adverts. Food adverts
contain a greater distortion in gender distributions for voice-overs, dominant
product users, and main character roles. Boys are presented as dominant
figures and role models.
Choate, 1972
USA
Content analysis
Adverts taken from 28 hours
of
children’s
Saturday
morning television
The content analysis reports that more than half of recorded adverts are for food
products (and 30% for toys). It also notes that 10% of adverts for vitamins
maintain that they are “sold to children in case you don’t eat right”.
A breakdown of advertising by food product category provided the following:
breakfast cereals (n = 92), candy, cookie and soft drinks (n = 51), snacks (n =
22), ‘drive-in’ (i.e. fast-food) restaurants (n = 20), main meals (n = 9),
pastries/puddings (n = 8), continental baking (n = 6) and soup (n = 2).
Condry, Bence &
Scheibe, 1988
USA
Longitudinal content
analysis
Adverts taken from 86.5 hours
of children’s television
Three content analyses produced longitudinal trends (N.B. this is better than
‘before’ or ‘after’, but random selection may negate this). Food products were
always the top advertised product during every year studied (65.4% for 1983,
48.4% for 1985 and 51.3% for 1987), though advertising is reported to have
declined due to increase of toy and other adverts. A breakdown of advertising for
different food product categories is provided for the years 1983, 1985, and 1987
respectively (in %): cereals, 31.8, 20.5 and 22.8; candy/snacks/sodas, 19.6,
14.5 and 17.2; restaurants 12.5, 8.8 and 7.9; other food/beverage, 1.5, 4.6 and
3.4.
Decline is most evident when compared with data from Barcus (1971a; 1971b):
total food advertising 82% (cereals 34%, candy/snacks 32%, restaurants 15%
A-20
and other food/beverages 1%). Seasonal variation finds that toys are the most
advertised product in December (46.6%) (approximately twice as much as the
other months: 23.2–26.4%), with corresponding fall in food adverts (cereals,
16.2% compared with 20.7% and 27.9% and candy/snacks, 7.7% compared with
20.5% and 24.2%). PSAs also declined (though many due to a decline in
general audience). PSAs for children and teens actually increased, but mainly
due to anti-drugs adverts (majority of 25 PSAs in 1987, by which time nutritional
PSAs had disappeared).
Connor, 2006
USA
Cross-sectional
content analysis
48 hours made up of four 4hour blocks (0900 to 1300
hours)
of
television
programming from 3 stations
aimed
at
toddlers
and
preschool children (PBS Kids
from a broadcast network,
and Playhouse Disney and
Nick Jr from basic cable
channels)
was
randomly
selected and recorded on
weekdays in May 2005.
In 96 half-hour blocks of preschool programming, the 3 stations had a total of
130 food-related adverts (1.354 food adverts per half-hour). More than one half
of all food adverts (76 of 130 adverts) were aimed specifically at children, and
the majority of those were for fast food chains (50 adverts) or sweetened cereals
(18 adverts). The primary advertising appeals used in food adverts aimed at
preschool children associated products with fun and happiness (82%) and/or
with excitement and energy (57%). Fast food adverts in particular seemed to
focus on building brand recognition and positive associations, through the use of
licensed characters, logos, and slogans. χ2 test analyses indicated that adverts
for fast food were more likely than adverts for other food products to use
animation (P=.004) and to incorporate licensed characters (P=.001) but were
significantly less likely to show food (P<.001).
Consumers
International,
1996
International
(inc
United
Kingdom)
Content analysis
Monitors television advertising
across 13 countries (11
European countries, the USA
and
Australia)
during
“children’s
programming”.
Each country aimed to
provide
20
hours
of
programming for analysis
although this varied across
the sample
The content analysis of television advertising by product category found that, in
all bar two channels surveyed (on Australian channel and one, of two, Swedish
channels) food was the most advertised product.
Of 13 countries, the United Kingdom had the most adverts in Europe during
“children’s programming (17 hours from 20 hours 10 minutes recording)”, but
this was behind the USA (24 hours) and Australia (29 hours). In the United
Kingdom this translated to 10 food adverts per hour (59% of all adverts); the
USA had 11 adverts per hour (45% of all adverts) and Australia had 13 adverts
per hour (39% of all adverts). The United Kingdom came second to the USA
among six countries in terms of number of adverts per 20 hours in 1996 (330
and 484 adverts respectively). The figure for France was similar (330 adverts),
the others being Germany (279 adverts), Netherlands (91 adverts) and Sweden
(37 adverts). In terms of proportion of food adverts, the United Kingdom was
second to the Netherlands (59% and 84% respectively) ahead of France (49%),
the USA (45%), Germany (41%) and Sweden (30%). That so many adverts are
for other products is said to indicate that adult viewers may be the target
audience during these advertisement breaks.
The United Kingdom and USA (1996) had highest numbers of food adverts
broadcast during children’s programming (n = 195 and n = 215 respectively)
A-21
compared with France (n = 166), Germany (n = 115), Netherlands (n = 77), and
Sweden (n = 11). The most advertised food category on United Kingdom
children’s television was confectionary (n = 55), which was also the most
commonly advertised category in the Netherlands and France. Next came
breakfast cereals (n = 32), which were the most commonly advertised in
Germany and the USA. Other categories of food product advertised on United
Kingdom television were ready-prepared foods (n = 30), restaurants (n = 23),
sauces & savoury products (n = 18), meat & meat products (n = 14), hot
beverages (n = 8) and all others (n = 15). The most advertised food products in
the United Kingdom were confectionary (55%), breakfast cereal (31%),
restaurants (18%), dairy products (17%), savoury snacks (15%), ready prepared
foods (8%), fish & fish products (7%), hot beverages (7%), soft drinks (6%),
cakes and biscuits (6%) and others (11%).
Across all countries confectionary, breakfast cereals and restaurants (virtually all
fast-food) accounted for more than half of all adverts. Confectionary alone
accounted for one fifth. This varied by country, with confectionary being
advertised in the United Kingdom approximately four times as much as in
Germany and 35 times as much as in Sweden. Confectionary also had the
greatest variety of manufacturers, though Nestlé, Cadburys and Mars were
frequent. Most (81%) of breakfast cereal adverts were for products high in sugar
or sodium (67%). The majority of cereals were manufactured by either Nestlé or
Kellogg’s. The third most advertised product was restaurants, such as
McDonald's. These adverts tended to promote the ‘experience’ or multi-item
meals (burgers and fries) rather than specific products and used brand building.
Again, these are products high in salt, fat and sugar, but low in fibre.
In contrast there were few adverts for fruit and vegetables. In the United
Kingdom the only advert recorded in this category was for frozen peas. Adverts
for fish & fish products tended to be for battered fish fingers or battered foods.
In this study only the United Kingdom carried out nutritional analysis (Sweden
attempted this, but had too few food adverts, Greece could find too little
information on food packets). This analysis involved examining high fat content
(> 30% energy) high sugar content (> 20% energy) and high sodium content (>
2.36 g/10MJ = > 6g salt). The analysis found that 62% of adverts were for
products high in fat, 50% for products high in sugar and 61% for products high in
sodium. In total, 95% of adverts were for products high in fat and/or sugar and/or
sodium.
The study also describes some creative strategies and reports on some
sponsorship (tie-ins) in the United Kingdom (other countries do not have
sponsorship or are not allowed; others have more sponsorship, especially
A-22
Australia). The tie-ins were for Potomus Park (Hipo Yogofrais, yoghurt) and for
the chart show (Twix, confectionary).
Consumers
International,
1999
International
Content analysis
Content analysis of 80 hours
of television in Hungary,
Poland,
Slovakia,
and
Slovenia
The content analysis found that food was the most commonly advertised product
(except during the run-up to Christmas and on the Cartoon Network where toy
adverts were more common. As with other research this Central European
study found that confectionary the most common product, together with
sweetened breakfast cereals, savoury snacks and soft-drinks, comprising over
three-quarters of all food adverts.
In terms of creative strategies, the study finds some evidence of sponsoring,
including chocolate (Poland, Slovakia, and Slovenia) and savoury snacks
(Poland). Some misleading adverts are noted, including ‘extra milk’ in chocolate
(Poland, Slovakia, and Slovenia) and chocolate giving energy (Slovenia). Sugarfree gum is also noted as being portrayed as an encouragement to eat
sweetened foods.
Inappropriate marketing is also discussed: adverts observed included chocolate
being kept in a medicine cabinet (Slovenia); a reference to being addicted to
savoury snacks (Slovenia); peer group exclusion (all countries). This last advert
involved three boys and a girl eating ice-cream, a fourth boy approaches and
gives her sugar-free gum to overcome the ‘wicked acids’, leaving the other boys
feeling excluded. Other inappropriate marketing observed included a cartoon
character used to promote yoghurt (Slovenia) and widespread use of free gifts
(eg. Kinder and McDonald's), competitions or children’s clubs (e.g. Lego). An
example of product placement is also provided in terms of a chocolate brand
seen on a television programme where St Nicolas gives out presents from a
basket on St Nicolas Day (Slovenia).
Though not part of the report, in-school marketing, direct mail, magazine and
internet advertising were also noted. For example, confectionary company
representatives handing out free samples and free gifts, with the company's logo
and a competition where children had to collect chocolate wrappers (Slovakia).
Consumers
Union, 1995
USA
Exploratory study of
in-school marketing
techniques.
The study includes the
analysis of posters, teaching
packs video, software, and
CD-ROMS. Of 111 teaching
guides examined, 77 were
learning kits, 29 sponsored
contests, and five reading
incentive programmes
A-23
All but one of 21 nutritional materials were reported as being sponsored by food
companies (e.g. Kellogg’s ‘Build on Good Nutrition’ – ‘Get going with breakfast’;
Mars’ ‘100% Smart Energy’).
McDonald’s programme avoids mentioning (its) fast-food: even in its ‘what’s on
your plate’, ‘balancing your act’ and ‘healthy growing up’, nutrition aids (video,
posters, teaching guide, booklets, student certificate, stationary, etc.) although
the logo is used or McDonald's is mentioned in the credits. The study reports
that this allows biased promotions which reflect inappropriate marketing on a
large scale.
Cotugna, 1988
USA
Content analysis
225 adverts taken from 12
hours of Saturday morning
television
The content analysis revealed that 71% of the 225 recorded adverts were for
food, and 80% of these were judged as having low nutritional value. The study
provides a breakdown of food advertising by product category: breakfasts
cereals (31%), cookies/candy/gum/pop-corn/snacks (34%), beverages (7%),
waffles/pastries (4%), canned pasta (5%), canned deserts/frozen dinners/driveins/peanut butter/oranges (13%), others including prepared meat and fish
products/diary product/oils/catsup (ketchup) (6%).
The study compares these data from 1987 with data from Brown (1976) (not
elsewhere included in this review) and Gussow (1972). The analysis finds little
change in nutritional content (previously, 76 and 84% respectively). Cereal
advertising was down at 31% (previously 38.5% and 41% respectively), but
offset by an increase in the ratio of sugared to non-sugared cereals (from 5:1 in
1976 to 12.5:1 in 1987).
Cowburn
Boxer, 2007
&
United
Kingdom
Cross-sectional
content analysis
15 magazines: 5 popular titles
(measured by the Youth
Target Group Index which
combines readership and
interviews data) for each of
the following age groups: <5,
6-10 and 11-19 years.
Includes all issues published
in November 2004 to January
2005 and June to August
2005. Internet food marketing
sites readers were directed to
in the magazines.
Food advertising appeared as (i) ‘cover mount’ free gifts and as (ii) part of the
main bound issue. The presence of free gifts varied with titles, and younger
children were offered free gifts more frequently than older children. All issues of
Pre-school magazines and two issues of Pre-teen/Teenage magazines
contained a free gift but none were food-related. 10% of issues of Early school
magazines had food based free gifts and all were confectionary. Pre-school
magazines contained no food advertising. Early school magazines contained
food 57 instances of food advertising: 48% for sugar-coated breakfast cereals or
sugary breakfast snacks, 24% for confectionary, 15% for sugared soft drinks and
3% for fruit and vegetables. Pre-teen/Teenage magazines contained 112
instances of food advertising: 27% for foods high in fat and sugar (mostly
confectionary and ice-cream adverts), 17% for pre-prepared dishes and cooking
ingredients, 12% for rice or cereal-based foods (most for healthier foods) and
11% for each of dairy and meat products.
54% food adverts in Early schools magazines referred readers to internet sites.
These were mostly from 2 titles but there was no consistency between the
quantity of adverts, referral to a website and the food product’s nutrient profile.
37% food adverts in Pre-teen/Teenage magazines referred readers to internet
sites. Most were from 1 title where 67% were for ‘less healthy’ foods. Food
adverts directed readers to 30 different websites, 83% of which were still active
at the time of analysis (6 months after last issue of sample magazine published).
44% websites were aimed at children, a further 16% contained sections for
children. 91% and 18% websites from Early school and Pre-teen/Teenage
magazines respectively were aimed at children. 64% websites contained games,
quizzes or tests and were more common websites from Early school (95%) than
A-24
Pre-teen/Teenage (36%) magazine referrals. 32% websites had a
‘spokescharacter’ with whom visitors could interact, most commonly from Early
school magazine referrals. 32% websites requested children seek parental
consent prior to use.
CWS Ltd, 2000
Dibb & Gordon,
2001
United
Kingdom
Content analysis
272 food and soft drinks
adverts were recorded from
broadcasting over a period of
37 hours 35 minutes in March
2000 in the United Kingdom.
The content analysis reported more food advertising on television during
children’s viewing time when compared with adult programming. It was found
that children are exposed to more food adverts than adults, with only 21% of
post-watershed adverts being for food products compared to 48% of adverts on
Saturday morning television and 58% of adverts during the children’s hour.
Of the adverts featured during children’s viewing times, 95–99% were for
products high in either fat (30– 40%), sugar (63–74%) or salt (27–49%).
Corresponding figures for broadcasts after 21.00 h were 88%, 25%, 25% and
49% respectively.
The National Food Guide recommended diet defines five products and how
much should be consumed of each. These are bread, other cereals & potatoes
(34%), fruit & vegetables (33%), milk & dairy products (15%), meat, fish &
alternatives (12%) and fatty & sugary foods (7%). Taking the three samples of
children’s viewing times, exposures to adverts for these categories of food were
bread, other cereals & potatoes (16%), fruit & vegetables (0%), milk & dairy
(10%), meat, fish & alternatives (4%) and fatty & sugary foods (70%).
An expert commentary on the creative strategies employed in food advertising
identified four emotional needs of children exploited by advertisers:
• the need for nurture and protection
• the need for stimulation
• the need for role models
• the need for peer group acceptance.
The consumer study reported that 73% of children ask their parents to buy after
seeing sweets and crisps advertised, and that 71% have bought something for a
free gift or token. If children are told ‘no’, a variety of 'pester-power' strategies
are used with only 19% of children reporting giving up and doing nothing.
Of adults participating in the study 77% wanted to see a ban on the advertising
of high sugar or fat products to children. On a four point rating scale, 68% felt a
free toy or gift was ‘very persuasive’, as did 65% with association with a
character. In comparison only 12% felt that claims about how healthy the
product is was ‘very persuasive’ with 40% rating this as ‘not at all persuasive’.
Combining scores of ‘very’ and ‘quite persuasive’ produced majority responses
for associations with cool, fashionable people (62%) and in-store display (68%).
A-25
Dibb & Castell,
1995
United
Kingdom
Content analysis
Two content analyses were
undertaken in the United
Kingdom during the weeks
11–17 June 1994 and 1–6
May 1995. Both weeks
involved 35 hours 20 minutes
of viewing. Observations were
made during ITV’s children’s
hour (8 hours 20 minutes),
Saturday morning television
(7 hours), the Big Breakfast
(10 hours) and late evening
(post-21.30-watershed) aimed
at adults (10 hours)
Food adverts constituted 7 out of 10 adverts on Children’s ITV (children's hour),
5 out of 10 for both Saturday morning television and the Big Breakfast (weekday
mornings), and only 2 out of 10 after the watershed. This is the inverse of the
proportion of adults who watch at each time.
In terms of nutritional assessment, adverts for food products high in fat, sugar or
salt reached 100% broadcast during children’s viewing times in 1994 and 98% in
1995. Advertising for such products during adult viewing time was 96 and 86%
respectively.
The national food guide recommended diet defines five products and how much
should be consumed of each. These are bread, other cereals & potatoes (34%),
fruit & vegetables (33%), milk & dairy products (15%), meat, fish & alternatives
(12%), and fatty & sugary foods (7%). The advertisements for bread, other
cereals & potatoes ranged between 16 and 40% across the different viewing
times. However, two-thirds of these were for breakfast cereals (mainly presweetened) and only 8% were for bread. Only 2 adverts of 549 were for fruit.
Between 0 and 5% of adverts were in the category of milk and dairy products
although none were for low-fat versions. Between 0 and 20% of adverts fell into
the meat, fish and alternatives categories. However, none were for fish or low-fat
products.
Advertising for fat and sugary products predominated, varying between 44 and
76% of all adverts observed (i.e. 6–10 times more than recommended
consumption of such products).
The study concluded that the foods we should eat least are the most advertised,
while the foods we should eat most are the least advertised.
Dibb, 1993
United
Kingdom
Review article
‘New’ data is provided from
recordings of two samples of
one week of children’s hour
television advertising
In terms of advertising spend, the review reports that in 1992 food and soft
drinks were advertised on television more than any other product sector (£523
million). Figures for only above-the-line advertising spend by food product
category are also provided: cereals (£83.9 million) chocolate confectionary,
excluding sugar confectionary (£72.7 million), soft drinks (£55.7 million) coffee
(£32.0 million), tea (£31.6 million), potato crisps and snacks (£22.6 million), icecream and lollies (£19.4 million), margarine & low fat spread (£18.8 million),
cheese (£16.9 million), fresh & frozen meat & poultry (£16.9 million), frozen
ready meals (£16.0 million), yoghurt (£15.4 million), milk & milk products (£14.7
million), stocks & stock cubes (£13.8 million), cooking sauces (£13.2 million),
frozen vegetables (£10.7 million) and butter (£8.5 million).
Excluding tea and coffee, the products with the highest spend are reported to
include brands heavily advertised to children. In contrast, advertising spend for
A-26
fresh fruit & vegetables was reported as only £4.5 million in 1991.
The review also mentions “below-the-line” marketing and gives examples of
other promotional channels including; comics (little used except for one-off
promotions), in-school promotion, computer games, sponsorship, packaging and
free gifts (toys and collectables).
The review also includes case study analyses of complaints about inappropriate
marketing:
• Milky Way – “you can eat between meals without ruining your appetite”
• Mars - “A Mars a day helps you work rest and play”
• Kellogg’s – reports on a misleading claim that a bowl of corn flakes with
semi-skimmed milk had less fat than a slice of brown toast with low-fat
spread.
• Farley’s bed timers – suggestion that a sugared drink had put a baby to
sleep.
• Lucozade – celebrity endorsement by Daley Thompson
The review also cites data from other content analyses of food promotion, e.g.
The Food Commission (1990) ‘Sweet Persuasion – A diet of junk food adverts’
which found that 53% of 92 adverts (from 4 hours of Saturday morning
television) were for food or soft drinks; eight times higher than any other
category or an average of 10 per hour. The review states that 78% were for presweetened cereals, sweets, crisps, fast-foods, soft drinks and other products
high in sugars, fats or both. In contrast, only 10% could be described as
encouraging a healthy diet. A follow-up study by the Food Commission (1992) of
the content of 190 adverts recorded from children’s television during one week in
May 1992 found that 47% of adverts were for food or soft drinks. The following
breakdown is provided by food product category: sweetened cereals (32%),
confectionary (16%), fast-food (10%), bagged snacks (6%), soft drinks (6%),
butter (6%), milk (6%), ice cream (4%) and others (14%).
The review also cites a content analysis from Young (1987) of 1750 adverts
shown after-school during 1983/84 which reported that 33% of adverts were for
food, and that 34% of these were for sugared products. Dibb concludes that the
frequency of adverts for such products is increasing in comparison to the
findings of Young.
Findings from Castell (1988) are also reported. This content analysis found that
62% of all food and drink adverts on TV-AM were for sugar-containing foods and
that 48% were highly sugared. Of these adverts, 80% were for soft drinks,
sweetened cereals, confectionary and ice cream.
A-27
Findings from Donkin (1992) are also reported. This study compares all adverts
on United Kingdom commercial television broadcast in the first six months of
1991 with child audience ratings. The study reports that such adverts are
“slanted towards snack/convenience foods” that are often high in added sugar
(50–55%), high fat products (45–50%) or both (25–30%). Adverts for fresh fruit
and vegetables were almost “non-existent” (apart from an advert for grapefruit).
Donkin’s (unpublished) compares audience figures obtained from the British
Audience Research Bureau (BARB) for 1 January to 1 July 1991 for the
percentage of total food advertising time and the percentage of total child
audience. For the Top 10 food product categories advertised during this time the
respective percentages are: chocolate (15.11%, 12.94%); breakfast cereals
(13.6%, 16.07%); frozen food (9%, 8.01%); prepared food (7.15%, 8.34%);
carbonated drinks (6%, 6.24%); crisps & savoury snacks (4.66%, 4.39%);
margarine & low fat spread (4.2%, 3.14%); biscuits & crisp-breads (3.92%,
3.97%), and restaurants & fast-food chains (3.41%, 5%).
Dickinson, 1997;
2000
United
Kingdom
Content analysis of
television
advertising
and
programming
Two weeks United Kingdom
television, 527 hours of
output, 872 programmes
Dickinson (1997) reports on a content analysis of 1186 food adverts by the
National Food Guide’s five categories: Bread/cereal/potatoes (26.9%),
fruit/vegetables (1.6%), dairy (9.4%), meat/fish/alternatives (15.4%) and
fatty/sugary foods (46.8%).
The study also reports on a content analysis of programmes (excludes 12.3%
unclassifiable): bread/cereal/potatoes (18.2%); fruit/vegetables (32.8%); dairy
(9.6%); meat/fish/alternatives (16.0%); fatty/sugary foods (24.3%).
The study found that the high-fat sugared diet commonly found in other studies
was present in the advertised diet, but not in the television programmes. The
programme diet mirrored the recommended diet (by National Food Guide)
regardless of whether these were fiction or non-fiction non-adverts.
Dickinson (2000) states that young people were particularly adept at recalling
the voice-overs in food adverts “almost verbatim.”
Dixon, Skully &
Parkinson, 2006
Australia
Observational
survey
Doolittle
&
Pepper, 1975
USA
Content analysis
257 individual checkouts in 24
randomly
selected
supermarkets situated within
a 20-kilometre radius of
Melbourne’s General Post
Office
A recording of Saturday
morning television on a single
day yielded 162 adverts
A-28
All supermarkets surveyed displayed food products at their checkouts, with most
checkouts displaying chocolate (87%), gum (81%) and sweets (80%). Only 7%
of checkouts had their display of foods or drinks out of the reach of children.
The content analysis of food adverts reported the following proportions: cereals
(40%); sweets & gum (18%); snacks & soft-drinks (12%); meal food (7%); other
food (7%). The remaining adverts were for toys (7%) and other non-food
products (8%).
An analysis of the created strategies utilized in the advertising found that free
gifts (toys) were promoted in 20% of adverts (all for cereals). Ambiance in the
advertising, categorized as ‘light’, predominates with only adverts for snacks and
toys below 90%.
In terms of the format of advertising, 71% of cereal adverts, 12% of sweet
adverts, 39% of adverts for snacks, 0% of adverts for meals, 0% of other food
adverts, and 0% of other product advertising had a mixed animation format.
An animation only format was found in 7% of cereal adverts, 44% of sweet
adverts, 50% of snack adverts, 50% of adverts for meals, 60% of adverts for
other foods, 0% of toy adverts, and )% of other product adverts. Characters
featured in the adverts were male dominated (84–94%) and 75% of adverts
used an announcer (this was not broken down by food product).
Eagle
&
Brennan, 2007
England
Cross-sectional
content analysis
30 leaflets, flyers, pamphlets
or menu-cards distributed in
one area of NW London over
st
4 months (1 December 2006
st
and 31 March 2007).
27 items (90%) related to foods whose advertising would potentially be restricted
if it were placed on TV (re. the early 2007 TV regulations). 74% for pizza, 10%
for Chinese food, and 3% each for Indian food and MacDonald’s. 3 leaflets
(10%) contained potentially healthy options and none offered healthy options as
part of their special promotions. None of the leaflets contained nutritional
information.
Egberts & Riley,
2004
Australia
Content analysis
75 hours of food advertising
on
Australian
television
directed at children and adults
Adult’s advertising contained more "core" food products (e.g. breads and
cereals, fruit and vegetables, and dairy). Children’s television advertising of food
used more cartoon (25.1%) and animated characters (13.7%), a faster pace (3fold) and the themes of magic, adventure and violence (50%) than found in
adult’s television advertising.
Elliott,
2008b
Canada
Cross-sectional
content analysis
367 ‘regular’ foods (dry
goods, dairy, produce and
frozen – excluding ‘junk’
foods) repacked to appeal to
children purchased from a top
(highest revenue and greatest
number of stores) Canadian
supermarket chain, Loblaws
in December 2005.
367 food products met at least 2 of the following criteria: direct claims to fun/play
on the pack; cartoons ‘pointedly’ aimed to children (i.e. not Jolly Green Giant,
the Pillsbury Dough Boy); tie-ins with children’s TV, merchandise or films;
foregrounding of unusual shapes, colours or tastes; puzzles or games targeted
at children.
2008a;
A-29
Of 367 products, 3 failed to provide information relating to the quantity of sugar,
3 to the salt and 2 to the fat contained in the product. 89% (326 products) of
those with nutritional data could be classed as PNQ (poor nutritional quality) due
to high sodium, fat or sugar, 61% ‘fun foods’ (222 products) were in dry goods
category, 17% (63) in frozen foods, 14% (53) in dairy and 1% fruit and
vegetables. Only small apples and baby carrots were specifically targeted at
children. 63% of ‘fun foods’ analysed made one or more nutrition claims on the
front of the box. Of the 326 ‘fun foods’ noted as PNQ, 202 (62.0%) had nutrition
claims on the front of the package. [NB: pre-packaged food in Canada requires
mandatory nutrition labelling.] 84% ‘fun foods’ used a cartoonish script or
crayoned font. 75% had a cartoon image on the front: an anthropomorphised
animal (48%) or a cartoon child (25%). 19% pictured the cartoon image engaged
in some type of physical activity.
Packages may appeal to fun by a direct reference to ‘fun’ on the box (7%), have
unusual product names or flavours (38%) or verbally emphasize the food’s
intractability (18%), collect points/enter a contest/free download (10%), display
an on-pack activity/game, interact with a website (11%), or provide a game
played with the food itself (3%).
Folta et al 2006
Federal
Trade
Commission,
2008
USA
USA
Cross-sectional
content analysis
Exploratory review
31 hours taped in 3 time slots
(0700-1000 and 1500-1800
hours
on
Tuesday,
Wednesday, Thursday and
0700-1300
hours
on
Saturday)
of
popular
children’s
television
programming (WB from a
broadcast
network,
and
Nickelodeon and The Cartoon
Network cable stations) was
randomly
selected
and
recorded during one week in
late September 2005.
There were a total of 987 advertisements and promotions in the 31 hours of
children’s programming, approx. 32 per hour.
An analysis of 2006 marketing
expenditures
(using
FTC
compulsory process orders)
and activities by 44 US food
and
drinks
companies
(primary marketers to youth
(2-17 years) in a range of
food categories). Analysed
media include: television,
radio, print, Internet activities
and other new electronic
media, packaging, in-store
Companies included: beverage manufacturers and bottlers; companies
producing packaged food (e.g. snacks, baked goods, cereals, and ready meals);
makers of candy and chilled desserts; dairy marketers; fruit and vegetable
growers; and quick-service restaurants (QSRs).
A-30
Excluding promotions, there were 711 adverts, 35% of which were for
food/restaurants. 27% unique foods ads were for breakfast cereals, 19% were
for restaurant meals, 13% for sweet snacks and desserts, 11% for both other
breakfast foods and juice drinks, 8% for dairy products and 3% for savoury
snacks. There were no ads for fruit, vegetables or carbonated beverages. Food
and beverage ads depicted children engaged in physical activity and associated
the advertised product with athletic ability significantly more than toy and game
ads.
In terms of persuasive techniques, foods being advertised were most often
associated with fun and good times (75%), pleasant taste (54.1%), being hip or
cool (43.2%), and feelings of happiness (43.2%). They were also associated
with toys giveaways, athletic ability, innovation or newness, friendship or social
success, magical or superhuman abilities, and deceiving or tricking adults.
The companies spent approx. $1,618.6 million to promote food and beverages
to children (2-11 years) and adolescents (12-17 years) in the US in 2006.
Carbonated drinks, QSR food and breakfast cereals accounted for 63% of this.
46% of the expenditure was on television advertising and 53% was by television,
radio and print combined. New media promotion accounted for only 5% of the
expenditure. Premiums (item free with food product or at discount price with
proof of purchase) and prizes totalled 4% expenditure – excluding toys
advertising,
event
sponsorship, and promotions
that take place in schools.
distributed with QSR meals. (When QSR toys included, it ranks second only to
TV advertising and is more than twice amount spent on marketing in any other
food category). Packaging and in-store display materials accounted for 12%
expenditure. Marketing in schools’ share was 11%, 90% of which was for drinks.
The final 15% expenditure covered: event and athletic sponsorships; celebrity
endorsement fees; cinema, video and videogame ads; product placements in
films, TV and videogames; cross-promotion licensing fees; and promotional
activities conducted in connection with philanthropic endeavours. 13% of all
reported youth marketing expenditure was for cross-promotions. For restaurant
food and fruits and vegetables, cross-promotions were nearly 50% of reported
child-directed (2-11years) expenditures.
Nature of promotions:
Integrated promotional campaigns with themes carried over the various media
used; cross promotions for films included TV, cinema and internet ads,
packaging and in-store displays, advergames, contests and sweepstakes and
premium items. Animated ‘spokescharacters’ appear across the different media
and make ‘live’ appearances.
Two-thirds of the 44 companies reported online youth-directive activities e.g.
space on website devoted to youth, independent websites for products that
appeal to children, advergames. Websites feature sports, music, free downloads
(screensavers, wallpaper, ringtones, MySpace layouts, activity sheets,
podcasts). Viral marketing such as e-cards/emails with links back to the site.
Peer-to-peer (not necessarily electronic) communications with young people
recruited to hand out promotional materials or samples.
Packing and point-of-sale materials used TV/film cross-promotions,
spokescharacters, premiums/prizes and staging of mini-events. Some
companies offered point systems via proof-of-purchase; redeeming merchandise
through the company’s website for example. Celebrity endorsers – actors,
athletes, singers, and musical groups – were featured in TV and print ads, on
the Internet and in store displays, primarily in ads directed to 8-14 years.
Sponsored events, performances and teams. Product placement occurred in a
few TV programmes popular with youth and some PG, PG-13 films.
Occasionally products integrated into videogame content. Marketing in schools
included point-of-sale displays at vending machines/cafeterias, sponsorship,
product samples, branded merchandise, nutritional/fitness materials.
Galcheva, Iotova
& Stratev, 2008
Bulgaria
Cross-sectional
content analysis
41.5 hours of children’s
television programming aired
between 0600 and 1830
hours on three national
A-31
Food/beverage advertisements accounted for 124 (33.4%) of all commercials,
with 96.8% being for unhealthy foods. 57% of them were aimed specifically at
children as the most advertised products were salty/sweetened snacks and
cereals, sweets, soft drinks/carbohydrate juices and salty foods, with no fruit or
Gamble
&
Cotugna, 1999
USA
Content analysis
networks (1 public Canal 1, 2
private NTV, BTC), were
videotaped on two weekdays,
Saturdays
and
Sundays
between 17 February and 2
March 2007.
vegetable commercials. Food advertisements used more themes of adventure,
animation, music and gifts to attract children’s attention, and gave information
based on the product’s taste, physical qualities, novelty, presence of
premiums/prizes. Of all food/beverage advertisements, 27.4% contained healthrelated information about the products; three-quarters of the advertisements
were shot with young normal-weight actors with a good/healthy appearance.
16 hours of Saturday morning
television yielded 353 adverts
The content analysis of food advertising showed that of the PSAs, 8 were
nutrition-related. Of the food adverts, 56.3% were for products in the ‘bread,
cereal, rice and pasta group’ (and, of this, most were in the least healthy
sugared cereals product category), 27% were for fast-food restaurants, 11%
were for fats, oils and sugars according to recommended diet food pyramid
definitions.
The study provides a detailed list of products. The top advertised food product
was high-sugar cereal (34.5%). Of food adverts 23.8% were for kids meals (e.g.
McDonald's), and the typical advertised kids meal consisted “cheeseburger,
French fries, soda and toy”.
The study also compared the food advertising trends from this 1996 sample with
four other data sweeps: Gussow (1972), Brown (1976), Cotugna (1988) and
Kotz & Story (1994). The comparison finds little nutritional change in adverts.
There were no adverts for fruit and vegetables.
The study finds a ratio of 19.5:1 sugared to non-sugared breakfast cereals
(those reported in other comparison studies range from 5:1 to 12.5:1).
Gantz
2007
et
al.,
USA
Cross-sectional
content analysis
1,638 hours (126 hours per
network) of programming from
13 networks: 6 commercial
broadcast (ABC, CBS, Fox,
NBX, WB and UPN), 6
commercial
cable
(ABC
Family, BET, The Cartoon
Network, Disney, MTV and
Nickelodeon) and one noncommercial
broadcast
network (PBS). Recordings
were made from 0600 to 0000
hours between May and
September 2005. Networks
chosen were the most popular
among 2-7, 8-12 and 13-17
A-32
This study covered advertising in all genres of programming viewed by children,
rather than just children’s shows. A total of 40,152 ads and 996 public service
announcements (PSAs) were identified and coded by type of product, service, or
issue. A total of 8,854 of these were for food or beverages.
Children aged 2-7 years see an average of 12 food ads per day on TV; 8-12
years see an average of 21 food ads per day on TV; and 13-17 years see 17
food ads per day on TV. 50% of all ad time on children’s shows was for food and
food is the largest products category among all the ads all ages saw.
34% TV ads targeting children and teens are candy and snacks, 28% for cereal,
10% for food, 4% for dairy products, 1% for fruit juices and none for fruit or
vegetables. 34% TV food ads targeting children and teens use taste appeals,
18% fun appeals, 16% appeal with the use of premiums/contests and 10%
appeal that the product is unique or new. 2% used nutrition or health claims as a
primary or secondary appeal in the ad and 5% used pep. 22% of the TV food
year olds from Neilsen data.
ads included a disclaimer (e.g., “part of a balanced diet”); 20% promoted a
website; 19% offered a premium; 15% portrayed an active lifestyle; 13%
included at least one specific health claim; 11% used a children’s TV or movie
character; and 7% featured a contest or sweepstakes.
Children aged 2–7 and 8-12 years saw an average of one PSA on fitness or
nutrition every 2-3 days and 13-17 year old saw less than one per week.
Guittard, 2008
France
Review
European published data
The research evidence regarding the effects of food promotion to children
extends over forty years. It is commonly known that the majority of foods
advertised to children are ‘highly processed’, i.e. high in fat and sugar and low in
nutrients. It is difficult to provide a global picture of the situation in the EU
because national self regulatory rules which govern TV advertising to children
may differ greatly. Data and figures are mostly available at national level.
Therefore one must be careful in trying to compare data between countries. In
the UK television advertising of HFSS products has experienced a significant
decline of almost 18% since 2003. This means that food and drink
manufacturers have voluntarily reduced their advertising expenditure on these
products by £42 million in the last three years, from £231 million down to just
under £190 million.
Gussow,
1973
USA
Content analysis
29 hours of Saturday and
Sunday morning television
yielded 388 adverts.
Of the 319 food adverts: 38.5% were for breakfast cereals (particularly
Kellogg’s), 17% were for “cookies, candy, gum, popcorn and other snacks”, 15%
were for vitamins, 8% were for beverages & beverage mixes, 7.5% were for
frozen waffles & pop-tarts, 5% were for canned pasta, with the remaining 9%
comprising adverts for canned desserts, frozen dinners, drive-ins, peanut butter
and oranges (Sunkist, termed as the “one positive note”).
1972;
In terms of inappropriate marketing, the study comments on adverts for vitamins
that are advertised with the slogan, “to keep you growing right even if you don’t
eat right” – calling these “offensive” (especially the advert for chocolate-covered
vitamins).
The study also states that television programmes are themselves counternutritional. For example, Sesame Street’s cookie monster is described as a
programme celebrity. The analysis observed no adverts for milk products
(except hot cocoa mix), and no adverts for eggs, meat, cheese, vegetables
and fruit. The study comments that a company (Libby) which sells fruit, meat and
vegetables, had advertising for a set of three “fun” frozen meals containing a
strangely imbalanced mix of high-carbohydrate foods that in the words on the
box “youngsters prefer”. The dinner comes complete with a packet of “chocolaty
super stuff” to add to your milk, the whole “seasoned and proportioned for the
younger tummy” whatever that means – and another example (Kraft) who sell
A-33
milk, cheese, yoghurt and ice-cream, yet whose only advertised product is
candy.
Hammond,
Wyllie
&
Casswell, 1997a
New Zealand
Content analysis
88 food adverts from the
children’s hours and primetime television.
A content analysis of the food advertising provided the following breakdown by
food product category: snacks (n = 36), fast-food (15), cereals (11), dairy (8),
pasta/rice (4), fruit/vegetables (6), drinks (3) others (5). The study also analysed
theme appeals in food advertising and provide the following breakdown:
acceptance (5%), achievement (2%), cool (3%), convenient (2%), desire (4%),
energy (4%), family (6%), fresh/quality (4%), fun (12%), health (6%), nationalism
(2%), natural (2%), power (3%), sex (4%), television (3%), solution (5%),
sophistication (4%), special offers (3%), sport (2%), taste (17%), value (4%) and
others (4%).
The study also provided a breakdown of theme appeals by product category:
snacks (taste, 18%; fun, 18%); fast-food (taste, 19%); cereals (health, 24%);
dairy (taste, 25%); pasta/rice (convenient, 21%); fruit/veg (taste, 14%; natural,
14%); drinks (energy/sport, 22%); Note: some of these numbers were very
small (e.g. nationalism or sport, 2% is five adverts).
The study suggests that the themes appealing to young people are those that
feature in advertising for low nutritional food products. The study concludes that
food advertising is not consistent with nutritional guidelines recommendations.
Hammond,
Wyllie
Casswell,
1997b; 1999
New Zealand
&
Content analysis
15
hours
of
recorded
children’s hour and prime time
television during weekdays
yielded 276 adverts
The content analysis provided a breakdown by food categories:
confectionary/snacks, 30%; fast-food, 17%; soft-drinks, 17%; breakfast cereals,
17%; dairy, 8%; pasta/bread, 4%; fruit/vegetables, 3%; retail food services, 3%;
others, 1%. Advertising for sweet snacks, fast-food services & restaurants,
drinks and breakfast cereals comprised 84% of all food adverts.
People-metre study: The people-metre data show how many of the 100
households with 17–19-year-olds are reached by food adverts and how often
they are viewed. These were multiplied together to provide an annual exposure
rate. When metering is accounted for cereals become less salient. At the top is
confectionary (68 adverts per 20 hours), followed by restaurants (52), drinks
(27), breakfast cereals (24), and dairy products (12). (No adverts were for cakes
& biscuits or fruit & vegetables).
When exposure rates were calculated from the above, sweet snacks were at the
top with 1121 advertising exposures plus 133 hours of promotions, from a total
of 1254 exposures; the highest product was chocolate (n = 512). Equivalents for
other categories were drinks (18.5% of the calculated exposure rate), fast-food &
restaurants (15%), breakfast cereals (14%), dairy products (4%) and
nuts/pulses/beans (4%). Taken together with all other categories (15.5%), the
A-34
total was 4298 exposures (3803 adverts and 495 promotions).
Comparisons were made with a 13-country study by Consumers International
(1996), and also Morton (1990) whose food categories where used in this study.
In these comparisons the United Kingdom had 54 adverts per 20 hours for
confectionary (second to Greece among the 13 counties), 32 adverts for
breakfast cereals (equal third) and 23 for restaurants (fourth). No United
Kingdom adverts per 20 hours were for cakes & biscuits, fruit & vegetables and
dairy products. New Zealand was third for food, top for confectionary, and
second for fast-food.
Harrison
&
Marske, 2005
USA
Content analysis
426
food
advertisements
aimed at general and child
audiences
Convenience/fast-foods and sweets comprised 83% of advertised foods.
Snacktime eating was depicted more often than breakfast, lunch, and dinner
combined. Apparent character body size was unreleased to eating behaviour. A
2000 calorie diet of foods in the general audience advertisements would exceed
recommended daily values of total fat, saturated fat and sodium. A similar diet of
foods in the child-audience adverts would exceed the sodium RDV and provide
171g (nearly 1 cup) of added sugar.
Harrison, 2006
USA
Cross-sectional
content analysis
40
hours
of
television
programming, rated as most
popular nationwide with 6-11
year olds, taped between
0700 and 2200 hours in northcentral Illinois over 5 weeks in
Spring 2003.
Analyses revealed notable differences between ads depicting Black characters
and ads without Black characters. Compared with ads without Black characters,
ads with Black characters were more likely to sell convenience foods, especially
fast foods, yet less likely to feature overweight characters. Ads without Black
characters, in contrast, were more likely to sell foods high in sugar.
Overall there were few differences in the nutritional breakdown of foods
advertised in ads with and without Black characters, mainly because both types
of ads tended to sell nutritionally unbalanced foods.
A 2,000-calorie diet of foods advertised in both types of ads would exceed
recommended daily values of total fat, saturated fat, and sodium, yet fail to
provide recommended daily values of fibre and certain vitamins and minerals.
Hawkes, 2002
USA
&
various
developing
countries
Exploratory review
Examines
the
literature
centering on four products –
Coke, Pepsi, McDonalds and
Yum! (KFC and Pizza Hut)
In terms of extent, the review reports that the main food brands are all in top 100
global advertising spenders and this advertising spend is increasing especially,
outside the USA. Promotional spend in the USA remained relatively stable
between 1994 and 2000 (US$ 2353 million to US$ 2347 million), but has greatly
increased elsewhere (US$ 1172 million to US$ 2211 million).
The review also examines the promotional channels being used to promote
foods and this includes television (“magic moment experience” rather than food
appeal), adverts in press or on signage, television programming and movie tieins, the Internet, point-of-sale, in-service marketing (including schools) and
A-35
sports (from Olympic to local) sponsorship. The creative strategies described
include ‘glocal’, novelty driven, premiums (free collectible-toys), children’s meals
(happy meals), children’s mascots, birthday parties, kids web sites (free Internet
access at point-of-sale), kids clubs with rewards for loyalty (i.e. brand building).
The review make reference to a McDonalds Snoopy promotion in South-East
Asia, where a toy doll dressed in a different cultural costume every day for 28
days was given away with the purchase of an Extra Value Meal. This caused
controversy when a paediatrician stated that if a child ate an extra value meal for
28 days it would gain a kilogram in weight. The promotion was also run in China,
where six meals had to be purchased before a toy was given away.
Television adverts are based on the following themes: family values, friendship,
local tradition, rebel/romance for teens, and fun and excitement for children (e.g.
cartoons).
Hill & Radimer,
1997
Australia
Content analysis
27 hours of television were
recorded
yielding
239
separate adverts for 275 food
products
The study also looked at PSAs, but of 29 PSAs, none were food-related.
The study examined a total of 239 adverts. To investigate the extent of food
advertising, a breakdown of advertising time was provided by food group: foods
high in fat and/or sugar, 48.0% (chocolate the most advertised within this
category at 24.9%); fast-foods, 28.5%; cereal/bread/rice/pasta, 9.1% (excludes
9.3% high in sugar, includes 4.8% low sugar breakfast cereals); fruit (including
canned), 5.9%; dairy products, 2.8%; vegetables, 0.8%; meat/fish/alternatives,
0%; others, 4.8%. To examine the nature of food advertising, an analysis of
disclaimers looked at messages consistent with dietary recommendations and
found that: 11.7% (of the total sample of adverts) contained messages to eat
more bread and cereals, 11.7% contained messages to eat foods containing
iron, 2.9% contained messages to eat a variety of nutritious foods, 1.7%
contained messages to decrease total fat intake, 1.7% contained messages to
decrease total sugar intake,, 0.8% contained messages to increase food and
vegetable intake, 0.8% to increase fibre intake, and 0.8% to limit salt intake.
None of the adverts contained messages about maintaining a healthy body
weight and eating foods containing calcium.
An analysis of theme appeals revealed the following percentages of food
advertisements with explicit and implicit consumer-related/promotional
messages (implicit messages in brackets): gifts, 20.1% (0); taste, 15.5%
(33.9%); fun, 14.2% (36.0%); cool, 13.4% (9.2%); popularity, 0 (4.6%);
competition, 2.9% (0). An analysis of nutritional appeals revealed the following
percentages of food advertisements with explicit nutritional messages: vitamins,
13.8%; minerals,11.7%; natural/pure/fresh, 11.3%; wholesome/goodness, 7.1%;
healthy/nutritious, 4.2% (with 16.7% implied); and breakfast is important, 2.9%
A-36
(0). The study also looked at additional foods within the item being promoted.
The top two were fruit, 9.6% (mainly in cereal advertising); and vegetables, 5.8%
(advertised in pizza or a sandwich).
The study was critical of the use of terms such as “natural” and “wholesome
goodness”, which tended to apply to chocolate products (27 of 38 and 14 of 17
respectively). The study concludes that only cereals are a healthily promoted
food meeting recommended nutritional guidelines and targets.
Ho & Len, 2008
Malaysia
Cross-sectional
content analysis
The marketing for 6 cereal
brands judged to be targeted
at children from the dominant
food companies in Malaysia
was studied over April-May
2008, including packaging, TV
commercials, websites and inschool promotions.
TV spots were recorded for 3
days (April 29, May 2-3 2008)
from 0600 to 2100 hours on
four free-to-air TV channels (3
private, 1 government owned)
where
children’s
TV
programmes
are
more
prevalent.
Holt et al., 2007
USA
Longitudinal content
analysis
All television national and
local-spot adverts (including
paid commercial ads, public
service announcements, and
promotions for a network's
own
or
affiliated
programming) aired during
the four weeks beginning
nd
th
Nov 2 2003, Feb 8 2004,
nd
th
May 2 2004 and Jul 4 2004
on all ad-supported networks
in the USA. Combined with
Nielsen audience estimates
data of exposure to 2-11 year
olds in 2004. Dates chosen to
match
similar
research
A-37
Three Kellogg’s and 3 Nestlé cereals chosen for analysis. Nature of the
marketing techniques in Malaysia included: On-pack promotions e.g. free gifts
and toys, movie tie-ins, puzzles and games; cartoon and animated characters;
health claims to reassure/make products attractive to parents and carers;
interactive websites; school programs; sponsorships of children’s sports
activities; and television advertising.
Extent of television advertising:
Of the 6 products, only Nestlé aired TV advertisements for its children’s
breakfast cereals. One product had two adverts, one featured frequently during
cartoon shows while another was normally aired during drama slots or late
evening programmes. The child-targeted advert featured an animated character
and taste appeals. The adult-targeted advert featured a mother and child,
nutrition appeals and school/sporting success of the child.
Children’s Overall Ad Exposure: 1977 and 2004
Using Adler et al’s (1977) National Science Foundation study data as a baseline,
2-11 year old children’s’ exposure to paid advertising fell by about 7% by 2004
and exposure to all advertising rose by about 17%. Children in 2004 spent less
time watching ad-supported television compared to 1977 and in 2004 ads were
shorter on average.
Exposure to Food Advertising: 1977 and 2004
Using Abel’s (1978) and Beales’s (1978) FTC studies as a baseline, children’s
food advertising exposure on TV in the USA has not risen and is likely to have
fallen modestly: 9% fewer food ads in 2004 than in 1977.
Changes in Exposure by Product Category
The mix of food ads (restaurants and fast foods, cereals, desserts and sweets,
snacks, sweetened drinks, dairy products, prepared entrees and other food)
seen by children in 2004 is more evenly spread across food categories than in
1977 (cereals and desserts and sweets dominated, with restaurant and fast food
and sweetened drinks also among the top categories).
published in 1978.
Sources of Children’s Ad Exposure in 1977 and 2004
Children got approximately half of their food ad exposure from children’s
programming (programs in which children are at least 50% of the audience) in
2004, compared to about one quarter in 1977. Saturday morning is a popular
viewing time for children, but children get almost as much advertising exposure
from one weekday’s primetime viewing (4.2% of the total) or from their Sunday
primetime viewing (4.1%) as from Saturday morning (4.3%).
The authors note that since the late 1970s, other marketing has likely changed
and new forms of marketing have emerged, including Internet-based advertising
techniques; these marketing activities are not covered this report.
Horgen, Choate
&
Brownell,
2001
USA/
International
Review article
Review
based
on
international literature
the
The review includes television advertising spend data and states that 24 of the
top 100 campaigns of the twentieth century were for food products. In 1997, US$
1.4 billion was spent on the promotion of food and food products in USA, with a
further US$ 1.2 billion on restaurants and drive-ins on network television, and
US$ 369 million on independent television.
The category of soft-drinks/snacks/confectionary was reported to have the fourth
biggest advertising spend at US$ 144 million. Restaurants and drive-ins were
the top advertised category on local television, with a reported US$ 1.3 billion
spent in 1997 (this is more than twice the next biggest category: auto dealers
with US$ 455 million). McDonald's moved from the fifth to the second biggest
USA advertiser between 1900 and 1992, and by 1997 was believed to be the
most prolific advertiser in Europe. The company is claimed to have stated that
40% of its adverts directly target children, and that the average child sees a
McDonald's advert almost every second day.
The review includes an analysis of in-school marketing, and states that eight
million children see USA in-school television Channel 1. This is a 10-minute
news programme with 2 minutes of adverts which provides schools with
televisions, video recorders and satellite dishes. The review cites figures that
state that 69% (31/45) of adverts featured on Channel 1 in a 4-week period were
for food, including gum, soft-drinks, fast-food, candy and snacks.
The review highlights the paucity of research in this area (e.g. analyses of
cafeterias with fast-food outlets or vending machines).
Howard, 2003
USA
Content analysis
100
randomly
selected
children’s
food
adverts
between 1987 and 1998
A-38
Results show a shift away from pro-social and healthful themes in the late 1980s
to anti-social and self-harming themes in the late 1990s. In 1998 food adverts
were dominated by social alienation, violence, crime and addiction.
Ji &
2001
McNeal,
USA/China
Content analysis
431 adverts recorded taken
from both Saturday morning
television and the children’s
hour
The content analysis found that 81.8% of Chinese adverts were for food,
compared with only 30.8% in USA where toys (55.5%) were the top advertised
product. The study reports that this is because of famine versus fun.
In the USA, the adverts were more fun-focused, in China more health-focused.
In the USA commercials were typically longer (25.5 versus 20.7 seconds).
There were also statistically more adult voiceovers (announcers) in the USA
(75.5 versus 62.9%) and also more speaking characters (37.1 versus 18.9%).
Some food-specific differences are reported as Chinese adverts were more
likely to show the product in use (71.8% versus 57.4% compared with 74.2 and
78.3% for all products). USA advertising featured fewer health appeals (7.0
versus 38.6%) and convenience appeals (2.0 and 6.1%), but more fun (43.5
versus 14.4%) and adventure (14.7 versus 3.0%) appeals than Chinese
advertising. Chinese adverts had a greater popularity appeal (4.0 versus 12.1% Confucian). However, Chinese products also more likely to give information on
quality (2.7 and 25.8%) and texture (1.0 and 14.4%). USA foods also more likely
to sell on uniqueness (18.2 versus 34.0%, compared with 12.4 and 14.1% for all
products).
In 10/132 Chinese commercials the brands featured were all non-Chinese, with
KFC, McDonald’s and Oreo cookies common to both China and USA.
Jones
Fabrianesi,
2006; 2008
&
Australia
Cross-sectional
intercept survey
100 adults aged 18-74 years
(M=37.6 years, sd 12.9), 62%
female, 64% with ≥1 child
recruited at a large regional
shopping centre in New South
Wales.
For three of products, 42–54% parents would buy the product after seeing the
child version magazine advert compared with 82–84% parents after seeing the
adult version magazine advert.
And for three of the products, 74–92% parents perceived that the adult version
of the magazine advertisement suggested the food was nutritionally beneficial
compared with 2–14% parents perceiving this for the child version.
The study results suggest that: (1) adults’ perceptions of advertised food
products appearing in both adults and children’s magazines and, most
importantly, purchase intentions for those products differ according to the
version of the advertisement seen and (2) adults clearly perceive distinctly
different messages in advertisements for the same products which are targeting
parents vs. those targeting children
Karupaiah et al.,
2008
Malaysia
Cross-sectional
content analysis
6 months of televised food
advertising from 6 out of 7 of
the
Nation’s
commercial
television networks.
Based on reported timings of children's programmes, prime time significantly
differed (p <0.05) between weekdays (mean = 1.89 +/- 0.18 hr) and weekends
(mean = 4.61 +/- 0.33 hr). The increased trend during weekends, school
vacation and Ramadhan was evident.
Over the six-month period, the mean number of food advertisements appearing
A-39
per month varied greatly between television stations (C = 1104; D = 643; F =
407; B = 327; A = 59; E = 47). Food advertising also increased the most in
September (n = 3158), followed by July (n = 2770), August (n = 2431), October
(n = 2291), November (n = 2245) and June (n = 2211). Content analysis of
advertisements indicated snacks were the highest (34.5%), followed by dairy
products (20.3%), sugars and candies (13.4%), biscuits (11.2%), fast food
(6.7%), breakfast cereal (6.4%), beverages (4.1%), supplements (0.9%), rice
(0.6%), noodles (0.5%), bread (0.3%), miscellaneous and processed foods
(0.2%).
The frequency of snack food advertised during children's prime time was 5 times
more than fast foods. The sodium content (mean = 620 mg per 100g) of these
snack foods was found to be highest.
Kelly
&
Chapman, 2007
Australia
Cross-sectional
content analysis
Sixteen
popular
(by
readership and circulation
data) Australian children's
magazines were selected and
back
copies
of
each
magazine were purchased for
publications
released
between
January
and
December 2006 (n = 76).
There were a high number of overall food references within the children's
magazines, with the majority of these being for unhealthy food products (63.7%
unhealthy versus 36.3% healthy foods, p < 0.001). The food groups with the
highest proportion of branded food references, and therefore paid marketing,
were ice cream and iced confection (85.6% branded references), fast food
restaurant meals (83.4%), high-sugar drinks (78.9%) and snack foods (73.4%).
Of all magazines, those targeting males and children aged 7-12 years had the
highest proportion of unhealthy food references (78.1 and 69.8% unhealthy food
references, respectively). Food references within children's magazines are
common and skewed towards unhealthy foods.
Kelly et al., 2007
Australia
Cross-sectional
content analysis
TV adverts from 357 hours of
television taped from all 3
commercial stations available
on
free-to-air
Australian
television in Sydney between
0600 and 2300 hours from
th
th
Sunday 14 to Saturday 20
May 2006.
Food advertisements occurred in similar proportions during children's viewing
hours and adult's viewing hours (25.5 vs. 26.9% of all advertisements,
respectively), although there was a higher rate of high-fat/high-sugar food
advertisements during children's viewing hours (49 vs. 39% of all food
advertisements, P < 0.001). There were even more advertisements for highfat/high-sugar foods during popular children's programmes, contributing to
65.9% of all food advertisements. Estimates of exposure indicate that children
aged 5-12 years were exposed to 96 food advertisements, including 63 highfat/high-sugar advertisements per week. Since 2002, there has been a reduction
in overall food and high-fat/high-sugar food advertisements.
Kelly
2008a
Australia
Cross-sectional
content analysis
119 food product websites
and 196 popular children’s
websites (selected on website
traffic data for 2-16 year old
Australian
children
and
previous
research
on
frequently marketed brands).
Websites contained a range of marketing features. On food product websites
these marketing features included branded education (79.0% of websites),
competitions (33.6%), promotional characters (35.3%), downloadable items
(35.3%), branded games (28.6%) and designated children's sections (21.8%).
Food references on popular children's websites were strongly skewed towards
unhealthy foods (60.8% v. 39.2% healthy food references; P<0.001), with three
times more branded food references for unhealthy foods. Branded food
et
al.,
A-40
references displayed similar marketing features to those identified on food
product websites e.g. graphics (78.0%), brand logos (75.0%) and corporate
logos (52.4%), links to the food product website (42.1) and inclusion of a product
package (35.4%).
Kelly
et
al.,
2008b; 2008c
Australia
Cross-sectional
content analysis
Advertisements broadcast on
all
three
commercial
Australian television channels
were
recorded
for
an
equivalent 1 week period in
May 2006 and May 2007 (714
h) in Sydney.
A total of 20,201 advertisements were recorded over the 14 days, 25.5% of
which were for food. Significantly more food advertisements broadcast during
children's peak viewing times (assessed from purchased commercial data),
compared to non-peak times, contained persuasive marketing methods of
promotional characters (e.g. celebrities or cartoon characters) (P < 0.05) and
premium offers (e.g. competitions, toys, rebates, vouchers) (P < 0.001).
During children’s peak viewing times, 61% of food advertisements were for high
fat/high sugar foods. During programs most popular with children, there were 3.3
non-core food advertisements per hour containing premium offers, compared to
0.2 per hour during programs most popular with adults. The majority of
advertisements containing persuasive marketing during all viewing periods were
for non-core foods. There were 18 times more food advertisements using
premium offers during programs popular with 5-12 year olds than during adults’
2
popular programs (χ =19.76, P<0.0001). Persuasive marketing techniques are
frequently used to advertise non-core foods to children, to promote children's
brand recognition and preference for advertised products.
Klebba, Stern &
Tseng, 1994
USA
Content analysis
135 adverts from 50 hours of
children’s programming
A content analysis of television advertising was undertaken by product category:
cereals (31.1%); toys (48.4%); restaurants (6.6%); candy (0.8%); other foods
(12.3%); other products (0.8%).
A content analysis of disclaimers in advertising by product category was also
undertaken: cereals (32.2%); toys (52.2%); restaurants (3.4%); candy (1.1%);
other foods (10.0%); other products (1.1%). These figures were examined using
Chi-square analysis, which demonstrated that, with the exception of breakfast
cereals, food advertisements were significantly less likely to have disclaimers
than non-food advertisements (i.e. toys) aimed at children. Cereals were also
significantly more likely (X2 = 10.63, P < 0.03) to use both audio and visual
disclaimers (31%) than toys (14%) or others combined (9%).
Time trend analyses were also performed with data from Stern & Haron (1984)
which revealed that use of disclaimers in children’s advertising has risen –
cereals from 88.5% to 97.7%; toys from 58.1% to 87.0%; restaurants from
10.7% to 75.0%; candy from 0 to 9.1%; other foods from 0 to 30.1%; other
products from 0 to 16.7%.
A-41
Kotz &
1994
Story,
USA
Content analysis
52.5 hours of Saturday
morning television yielding a
total of 997 adverts
The content analysis of 564 food adverts classified 46.3% as ‘fats, oils & sweet
food’ group in the USA-recommended diet food pyramid. The product groups
‘bread, cereal, rice & pasta’ were represented in 37.5% of adverts, fast-food
restaurants in 10.8% of adverts (McDonalds, Burger King and Pizza Hut), ‘milk
cheese & yogurt’ in 4.5% of adverts, and ‘meat, poultry, fish, dry beans, eggs &
nuts’ represented in 1.8% of adverts. There were no adverts for fruit or
vegetables.
By individual products, sugared cereals were the most advertised product
(23.0%), followed by candy (15.0%), cereal with sugar as the main ingredient
(10.3% – recorded in ‘fats oils and sugar food group’, rather than the ‘bread,
cereal, rice and pasta’ group), low sugared cereals (6.0% – less than 20% sugar
by weight) and soft-drinks (5.6%).
When examining creative strategies, authors rated whether each advert
contained explicit or implicit messages (N.B. levels of viewer reliability on each
measure are reported). This process rated 2.4% of food adverts as containing
an explicit “healthful & nutritious” message. This is behind “taste” (36.2%), “free
toy” (16.9%), “fun” (16.7%) and “cool or hip” (7.3%). However, the authors rate
49.1% of adverts as containing implicit “healthful & nutritious” messages, the top
implicit message. Most commonly resulting from “a complete/balanced/nutritious
breakfast claim”. Other implicit message ratings being “taste” (35.9%), “fun”
(29.1%), “cool or hip” (10.3%) and “convenient” (7.3%). The study concludes the
diet of the Saturday morning pyramid is the “antithesis” of the recommended
diet.
Kunkel & Gantz,
1992
USA
Content analysis
604 hours children’s television
were recorded from Saturday
morning
television,
the
children’s hour, and weekday
morning television yielding 10
325 adverts
The content analysis finds significant differences across television networks, by
the number of adverts: more feature on network broadcast television and fewest
observed on cable channels). Differences are also observed in terms of the
diversity of products. The authors state that this may be due to lower costs or
lack of regulation. The study cites an (unethical) phone-line advert.
The content analysis provides a breakdown by product category by all three
media: For major television channels: toys (17.2%); cereals (31.2%); snacks
and drinks (32.4%); fast-food (8.7%); healthy food (4.6%); other (5.8%). For
local television channels: toys (42.1%); cereals (22.7%); snacks and drinks
(15.6%); fast food (5.6%); healthy food (1.7%); other (12.4%). For cable
television channels: toys (24.7%); cereals (15.9%); snacks and drinks (15.8%);
fast food (3.8%); healthy food (4.3%) other (35.5%). Toys comprised 33.8% of
adverts across all media, cereals and breakfast 22.4%; snacks and drinks,
18.4%; fast food, 5.7%; healthy food, 2.8%; and other, 16.9%.
Same food products dominate, but especially in broadcast networks, toys more
A-42
on independents, other products remarkably more on cable.
The study also reports on an analysis of theme appeals in advertising in terms of
fun, flavour, performance, social context, power, appearance, personal gain,
texture and other. These are broken down respectively into percentages by
product category: toys, 25.3, 1.6, 37.6, 3.6, 19.7, 4.9, 0.2, 0.1, 7.1; cereals,
15.4, 46.6, 2.6, 6.9, 0.6, 7.0, 1.8, 7.3, 11.9; snacks & breakfast, 26.1, 36.8, 4.4,
15.7, 0.5, 4.0, 1.7, 2.0, 9.0; fast-food, 71.9, 3.8, 2.3, 13.2, 0, 0, 1.9, 0.2, 9.0;
healthy food, 46.7, 15.7, 4.1, 6.1, 0.5, 4.6, 3.0, 0, 19.2; other, 27.5, 1.2, 23.7,
6.7, 3.7, 2.3, 17.9, 0.1, 16.9; total, 26.6, 18.8, 18.3, 7.7, 7.6, 4.6, 4.0, 2.1 and
10.4%.
The study also reports on the use of disclaimers in adverting. More than half of
all advertisements featured disclaimers (n = 6195). Percentage breakdowns in
terms of the proportion of audio, video and both audio and visual disclaimers
respectively are provided by product category: toys, 33.4, 47.3, 19.6; cereals,
87.3, 4.6, 8.1; snacks & drinks, 17.7, 68.9, 13.4; fast-foods, 14.1, 58.7, 27.2;
healthy food, 42.9, 57.1, 0; other, 16.2, 31.6, 52.2; total, 49.6, 32.1 and 18.3%.
Kuribayashi,
Roberts
&
Johnson, 2001
USA
Content analysis
Recordings
of
Saturday
morning television and primetime television (20.00–21.00
h) in mid-September 1997
yielded 145 and 136 adverts
respectively
27.0% of adverts during Saturday morning television, and 16.8% of adverts
during prime-time television were for food products. Non-significant differences
were reported in the number of adverts featured during children’s programming
and prime-time television. Adverts were more likely to feature food products on
during Saturday morning television (27% compared with 16.8% of prime-time
advertising).
In terms of nutritional assessment: products high in fat featured in 50.0% of
Saturday morning television and 65.2% of prime-time television; products high in
sodium featured in 50.0% of Saturday morning television advertising and, 69.2%
of prime-time television advertising; products high in cholesterol featured in 25%
of Saturday morning television advertising and 17.4% of prime-time television
advertising; products high in sugar featured in 50.0% of Saturday morning
television advertising and 13.0% of prime-time television advertising. The total of
adverts classified as featuring unhealthy products were 97.5% and 78.3%
respectively. A t-test analysis found that products advertised during morning
television were significantly more likely to be classified as unhealthy (overall)
and high in cholesterol or sugar.
Larson, 2003
USA
Content analysis
595 adverts featuring child
characters
Lemos, 2004
Portugal
Content analysis
Adverts from 84 hours of
programming
A-43
Results show that over one-third of the adverts that featured children contained
aggression. More than half of the aggression incidents occurred in adverts that
featured only white children.
In the 504 advertisements: 26% were for breads and cereals (sugared cereals);
35% for sweets (chocolates and cookies); 12% for soft drinks; 21% for dairy
products; and 6% miscellaneous. This pattern distorts the Portuguese food
guide recommendations as it has much more fats and sugars and less fruits and
vegetables.
Lewis
1998
&
Hill,
United
Kingdom
Content analysis of
children’s television
advertising
91.33 hours of children’s
broadcasting yielded 828
adverts for analysis
The content analysis provides a breakdown of advertising by food product:
cereal, 30.1%; confectionary/savoury snacks, 29.8%; other food, 34.3%; and
fast-food, 5.8%. Comparisons are made with Dibb & Castell (1995), finding a
reduction in food advertising from 62.8% to 49.4%, but still remaining the top
advertised product category. Top foodstuffs (60%) being confectionary, cereals
and savoury snacks. Some evidence of a trend from fast-food to convenience
foods (sauces, ready-meals etc.) which tended to be broadcast at tea-time
(maybe seasonal, more toys in this study).
The study also compares satellite and terrestrial television, finding more
breakfast cereals on satellite television. It is difficult to partial out effects of time,
season and channel. In terms of creative strategies, food adverts were reported
as being significantly more likely to be cartoons, to use humour, to feature a
story and to be mood-altering/fun in nature, but less likely to make claims of
value for money. Food adverts more than toys, had magical/fantasy theme, as
did child-oriented adverts, which differed greatly across all products in
comparison to food adverts only. Authors suggest this difference was less with
food adverts so as child and parent watching together would influence decision
to purchase.
Lobstein et al.,
2008
32 countries
6
Cross-sectional
content analysis
13 breakfast cereal products
marketed as being suitable for
children on sale in many
(between 7 and 25) of the 32
countries during April 2008.
Recordings
of
prime-time
programmes from
the 32 countries
week in April 2008.
children’s
television
nearly all
during a
6
The nutritional quality of breakfast cereals marketed to children: Of the 13
products examined, only one had ‘medium’ levels, while all remaining products
had more than 25% sugar many had sugar levels above 40%, and in one
sample the sugar level was 55%. Salt was added to nearly all products, typically
at around 1% of the product’s total weight but ranging up to 2.5%. Nearly half
the products could be classed as ‘low fat’ and all the remaining ones as ‘medium
fat’. The nutritional quality of products with the same name and similar
appearance varied considerably between different countries.
On pack marketing tactics: Examples of methods used to make products
attractive to children: characters from TV and website tie-ins, puzzles, gifts,
appealing/fantasy shapes, coloured ingredients, added flavour agents, movie tieins, sports equipment vouchers, clubs/societies. Examples of methods used to
make products attractive to parents: Whole grain, high fibre, low fat; helps
Europe: Belgium; Czech Republic; Denmark; France; Germany; Ireland; Italy; Netherlands; Norway; Poland; Portugal; Slovenia; Spain; Sweden; Switzerland; the United
Kingdom.
Asia and Pacific: Australia; Fiji Islands; China, Hong Kong Special Administrative Region (SAR); India; Indonesia; Malaysia; New Zealand; Russia; Republic of Korea;
Thailand. Americas: the USA; Argentina; Brazil; Chile; Peru. Africa: Kenya.
A-44
education attainment; front-of-pack nutrition claims; vitamin claims; fat free, low
fat;
energy;
lowers
cholesterol;
natural
ingredients;
no
added
colours/flavours/preservatives; good for child growth/teeth/bones; happy
families; approved by parents.
Television advertising: Examples include cartoon characters, fantasy
adventure, prizes, free gifts, sports celebrities, mother-and-child interactions.
Websites: Packs led purchasers to seek more nutritional information by visiting
websites, voucher and coupon give-aways on packs (such as swimming pool
entry vouchers) were further explained on websites, and most significantly,
children were encouraged to visit company websites for entertainment, including
games, puzzles, cartoons, ringtones, and other rewards. Company websites did
not restrict access/require parental approval to prevent young children gaining
entry.
TV adverts have been used for many years, and in some countries the use of TV
advertising is beginning to decline in favour of other approaches, including
Internet sites and direct marketing. Note that sweetened breakfast cereals are
being promoted to children through other activities, such as school classrooms
and teaching activities, and sports sponsorships for children’s junior
championships.
Longman, 2000
United
Kingdom/
international
Review article
The
review
uses
a
convenience
sample,
an
industry Opinion Survey by
Datamonitor
and
reports
figures
from
Consumers
International (1996)
The Industry Survey by Datamonitor identifies future trends in marketing to
children (as more they experience more economic independence due to more
money and family structure change). Targeting tools that are described as likely
to decrease in use in the future are licensed characters and in/on pack
promotions. Those reported as likely to increase in the future are healthy
products and fun food. In terms of fun tools, the use of characters, in/on pack
promotions and collectible toys are likely to decline in use in the future, and .use
of interaction, play with food, and competitions is likely to increase or remain the
same.
The following creative strategies are identified for use with the different groups:
• children – shapes, cartoons (novelty and pleasure), no health
• adolescents – cool style, some health
• adults – convenience, much health
• family (biscuits) – combining.
Bipartite relationships – child as independent buyer, able to exert influence over
other buyers for products for themselves or influence other buyers for products
for the buyer; and as a future buyer.
A-45
Case studies (in terms of appeals in promotions) are also described:
• Kraft Lunchables – convenient for mum
• Quaker Dinosaur Eggs – fun for kids, health for mum
• Yoplait Frubes – fun for kids, convenient and healthy
Each of above undergoes a SWOT analysis, focusing on the interactions
between health/convenience/pleasure. Statistics on product launches are also
provided by Datamonitor: Confectionary, 38.2%; bakery/cereals, 22.1%; dairy,
10.6%; soft-drinks, 9.1%; desserts, 5.8%; meats 3.5%, snacks, 3.5%; ready
meals, 3.2%; canned food, 2.1%; others, 2.0%.
Longman, 2002
United
Kingdom
/
International
Exploratory review
Convenience sample and
Industry Opinion Survey by
Datamonitor
The report highlights the potential growth in online marketing, by examining
various examples of children’s food and drink products’ web sites.
Reports that the United Kingdom youth (after USA youth) spend more time
online than youths in Germany, Sweden, Netherlands, France, Italy and Spain).
The report also describes some of the creative strategies being used by the food
industry to market its products to young people over the Internet. Apart from
direct selling, strategies included the use of cartoon styles, music, quizzes,
competitions, games, animations, educational content and links to other sites.
These activities can be complemented by offline advertising and promotion, to
help to build a strong consumer-brand relationship, and the incorporation of
parent/teacher-friendly material to ensure the child is not discouraged from
accessing the web site and links from other sites. In short, the site itself needs to
be advertised and must also provide inducements so that it will be accessed in
the first place.
As well as helping to build a strong consumer relationship with young people,
Internet advertising allows many more specific advantages of to the sellers.
These include the addictive component of web-surfing, particularly games or
other features which may foster repeat visits to the site. Sites having an
educational content can exploit the use of the Internet in schools (in-school
marketing), and sites which encourage registration or participation in
competitions can help firms collect consumer/market data to facilitate direct
marketing (e.g. a customer account).
The report presents some case studies of food and drinks promotions over the
Internet. Examples are shown of promotional web sites for companies selling:
• United Biscuits BN – use of animation and brand imagery
• Eden Vale Munch Bunch – educational emphasis
• Nestlé Nesquik – use of games to encourage return visits
• Petits-Filous Frubes – use of animation, brand imagery and games
A-46
•
•
Ferrero Kinder Surprise – parent-friendly design
Tango soft-drink – use of games, registration and soliciting for other
products.
Macklin & Kolbe,
1984
USA
Content analysis
A total of 144 adverts were
obtained for analysis via three
consecutive
Saturday
mornings on three major
television
networks.
Repetitions were excluded,
yielding
64
unique
commercials
Of the 64 adverts, more than two-thirds (69%) were for food. The sample differs
from others which have only looked at toys however the study still concludes
that there are stereotyped sex roles in children’s adverts.
Maher
2006
USA
Longitudinal content
analysis
Television advertising from
Nickelodeon and WB which
air
the
top
children’s
programmes (Nielsen data)
taped in the morning and after
school for five days, Saturday
morning, and Sunday morning
during 1 week in March 2000
and the same week in 2005.
14
hours
of
children’s
programming yielded 155
commercials
In 2000, 45 executions consisted of adverts for food, beverage, candy or
restaurants (excluding duplicates PSAs, adult commercials and network
promotions). In 2005, 80 food adverts were recorded (same exclusions). Results
indicate that food processors and restaurants have not changed their advertising
messages to children in response to the multitude of pressures the industry is
facing. Specifically, this pre-post longitudinal comparison shows no significant
change regarding types of food products advertised and type of appeals used in
the ads directed to children.
Data from post-1990 reports,
studies, surveys and projects
on-going at the end of 2004
were gathered from academic
journals,
government
departments,
commercial,
pubic interest and market
research organizations were
collected
May
2004
–
February 2005.
The data showed that food advertisements during children's TV overwhelmingly
promoted ‘unhealthy’ foods, with very little promotion of fruit and vegetables and
other ‘healthy’ foods. The extent of unhealthy food marketing to children varied
between countries with estimates ranging from 49% in Italy to nearly 100% in
Denmark and the UK.
et
al.,
Maryam et al.,
2005
Islamic
Republic
Iran
Content analysis
Matthews, 2008
Austria;
Belgium;
Czech
Republic;
Denmark;
Estonia;
Finland;
France;
Germany;
Greece;
Hungary;
Iceland;
Ireland; Italy;
the
Netherlands;
Norway;
of
Review:
questionnaire
data-collection
protocol
and
“Ashimashi’s” puffed cereals were the largest category of advertised foods
(36%). The message most frequently used to promote the sale of a product was
“taste” (56%). The most frequent appeal of food adverts was “attributed quality”
(67%). Half of the nutritional messages were scientifically untrue or misleading.
In the UK, children viewed an average of 5 TV adverts per day for Core
Category products (food, soft drinks and chain restaurants), the vast majority of
which were for food items considered to be unhealthy in having a nutritional
content high in fat, salt or sugar.
Across countries, commonly used creative strategies used when targeting
children included: referencing movies and their characters, using child-related
appeals to play, fun, action-adventure, humour, magic or fantasy; using cartoon
or celebrity characters.
A-47
Portugal;
Slovenia;
Spain;
Sweden
&
the
United
Kingdom
Recent UK statistics suggest that though spending on advertising in the food
sector may be increasing, the proportion spent on TV adverts may be declining.
Nonetheless, across countries where data was reported, the vast majority of
food promotion was through TV, with food promotion through radio, magazines
and cinemas taking a low and possibly declining proportion of advertising
spending.
Schools, on the other hand, represented a growing marketing channel and the
Internet was also a new and growing medium, where creative strategies included
cartoon-style games, competitions, educational materials and links to food
company websites.
Mazur
2008
et
al.,
Poland
Cross-sectional
survey
All 44 primary (N=8971
pupils)
and
secondary
schools (N=5225 pupils) in
Rzeszow, Poland.
Shop owners proximal to the
schools description of food
purchases by students one
week before school visit.
Recommended foods like milk, yoghurts and fruit were offered by only 40·9% of
shops.
There was a correlation between foods offered in the shop and foods purchased
by students. In schools, 40·9% (95% CI 25·8, 56·0%) of shop windows displayed
or advertised ‘healthy’ foods while 9·1% (95% CI 0·0, 17·9%) of shops displayed
advertisements of food companies. The difference between display of ‘healthy’
food in shop windows and display of food on company advertisements was
2
significant (likelihood ratio χ test, P < 0·04). Type of school (primary v.
secondary) was not significant factor in advertising or purchasing pattern.
School’s head-teacher or
teacher’s
description
of
school policy on food ads and
sponsorship.
Messner et al.,
1999
USA
Content analysis
Advertising featured during
sports programming during
one week yielded 722 adverts
for analysis
A content analysis of 722 adverts finds 11% are for snacks/fast-food (highest
category was automobile adverts, 20.5%). The highest level of snacks/fast-foods
advertising was found during wrestling (21% – just behind automobiles) which
was not regarded as a real sport by the authors, and this was the highest
product category (14%) with extreme sports.
An analysis of creative strategies revealed that some sponsorship (e.g. “scores
brought to you by…”) and adverts or corporate banners were visible from the
field of play or logos on equipment or opening shots [N.B this is not broken down
by food category]. Some celebrity endorsement and tie-ins with branding (e.g.
baseball superstar hits into big Mac land – McDonald’s) are also reported.
Molnar
2008
et
al.,
USA
Cross-sectional
study by telephone
313 primary (elementary and
middle) school officials from a
stratified random sample of
US public schools reporting
A-48
According to school officials, 37.7% of primary schools nationwide participated in
fundraising, 31.6% participated in incentive programs, and 16.3% participated in
exclusive agreements with a corporation that sells FHFS or FMNV.
Moore,
2006;
Moore
&
Rideout, 2007;
Moore, 2008
USA
Cross-sectional
content analysis
their school’s participation in
marketing
activities
with
corporations that sell foods
high in fat and sugar (FHFS)
and
foods
of
minimal
nutritional value (FMNV)
87.5% of school officials reported that their schools would not be forced to
reduce programs if marketing was prohibited, and 53.7% supported the
increased regulation of FHFS and FMNV marketing.
Websites of 77 major food
advertisers, who advertise
heavily to children (chosen by
studying TV listings and
competitive media reports),
during the summer and fall of
2005.
85% leading food brands that target children on TV in the US are also either
directly targeting children on the internet or provide online content likely to be of
interest to children.
68% sites directly centred on young children (2-11 years); 12% emphasized
information for adults or parents, but child-oriented content was also embedded
(often in a separate section); 9% seemed to target teens directly, but there were
activities of interest to a younger audience as well; 11% contained content for
each age group. 50% included products had an on-pack website address.
73% of websites studied included advergames†; 79% of websites included a
‘brand benefit’ claim on their site; 64% used viral marketing and this was more
common on sites for children and teens than those including adult content; 53%
websites had TV commercials to view; 25% offered users membership, 13%
required parental permission for this; 13% websites included polls/quizzes which
could be used for market research; 47% websites had a film or TV tie-in; 38%
used incentives for product purchases; 73% websites allowed customisation,
enhancing the user’s involvement; 76% websites provided brand-related content
the child could use when they left the site e.g. wallpaper or screensavers; 72%
websites studied contained nutritional information and 27% had information
about a healthy diet; 35% websites had educational content; 97% of sites
included information specifically labelled for parents.
The researchers highlighted 11 issues with public policy implications:
1. Common Online Marketing Practices: Unhealthful brand nutritional profiles;
Persuasion potential of “advergames”; Ethics of viral marketing; No limits on
advertising exposure; Limited use of “ad breaks”; Opportunities for corporate
research abound; Information for parents is available.
2. Varied Online Practices: Children’s online privacy protections; Direct
inducements to purchase; A new venue for licensing and host selling; Learning
potentials and pitfalls.
†
An internet-based or downloadable videogame advertising a brand-name
product by featuring it as part of the game.
A-49
Morgan
2008
et
al.,
United
Kingdom
Cross-sectional
content analysis
503 hours of television videorecorded from the 4 most
popular
British
children’s
commercial
television
channels
(free-to-air
ITV
England/ Wales and CITV
and subscription satellite Nick
Junior and Nickelodeon) at
the most popular viewing
times for children over 5 oneweek periods from October
2006 to January 2007.
Analysis of the recordings revealed that 16·4% of advertising time was devoted
to food products; 6·3% of all advertising time was devoted to high-sugar and
potentially cariogenic products (those which readily give rise to tooth decay).
Sugared cereals were the most commonly advertised high-sugar product,
2
followed by sweetened dairy products and confectionery (χ =6524·8, df=4,
P<0·001). The advertising of confectionery and high-sugar foods appeared to be
2
influenced by school Christmas holidays (χ =69.7, df=8, P<0·001).
Morton & Moore,
2004
Australia
Content analysis
75
hours
of
television
programmes
broadcast
between 16.00 and 21.00 on
three
South
Australian
commercial TV stations in
June 2003
Over half of the adverts were for convenience foods. Many explicitly devalued
home cooking. The adverts verbally and visually denigrate home cooking and
show its rejection; it is portrayed as an unrewarding use of women’s time.
Convenience foods are presented are compatible with “good mothering”.
Morton, 1984
Australia
Content analysis
120
food
advertisements
recorded between the hours
of 16.00–18.00 everyday
(16.00–17.00
being
the
children’s hour and 17.00–
18.00
being
prime-time
television) for five days
(weekdays only) during one
week in June 1984
30 hours of television yielded only 120 food advertisements (which averages 4
per hour and 2.5 per children’s hour).
Morton, 1990
Australia
Content analysis
In total, 45 hours of television
were recorded over 5 days
(weekdays only) from 3 hours
per day (16.00–18.00 and
19.00–20.00) during 1 week
in April 1989, yielding 851
A-50
One channel (ADS-7) had no food advertisements during the children’s hour,
another (SAS-10) had six food advertisements and the third (NSW-9) had 32.
Less difference between the frequencies of food advertisements was observed
in the following hour (25, 26 and 31 food adverts respectively).
Food advertisements were also analysed by food category. In total, only eight
adverts were for breakfast cereals (all were for unsweetened cereals and none
featured during the children’s hour) compared with 44 adverts for confectionary,
18 adverts for pies/pasties, 10 adverts for fast food and 10 adverts for snacks,
and only two adverts for soft drinks. Two adverts were for fresh fruit (apples), but
these were then seen being baked into high-energy foods. The study concludes
that although there is less food advertising to children in South Australia than
elsewhere, it is still unhealthy.
A content analysis of television advertising revealed that food was the top
advertised product category with 412 advertisements (48%). The same four food
product types were found to predominate as elsewhere: 19.9% for “chocolate &
confectionary”; 17.4% for “food services & restaurants”; 17.2% for “breakfast
cereals”; and 17.0% for drinks & drink mixes” throughout all three television time
slots. Although this was true of all time slots, products with less appeal to
advertisements
minutes
lasting
417
children (e.g. rice or margarine – especially low-fat, rather than butter) featured
less during the children’s hour (i.e. between 16.00 and 17.00).
The content analysis also recorded fewer advertisements more generally during
the children’s hour (87.3 minutes) compared with the hour immediately following
the children’s hour (168.15 minutes) and the first hour of “peak-time” viewing
(161.15 minutes). However, the number of food advertisements recorded within
each time slot was similar (130, 156, and 126 respectively). Therefore the
percentage of food advertisements in the children’s hour was more than that of
the hour immediately following the children’s hour (46%) and the first hour of
“peak time viewing” (37%). Only 50 advertisements were for low-processed
foods, and very few advertisements for such products were recorded during the
children’s hour.
However, the study notes that on one channel in the children’s hour 93% of
adverts were for food and that one programme that featured during the hour was
sponsored by Kellogg’s and McDonald’s. Also observed, are differences in the
promotion of food products across the different time slots. For example, the
study reports that 32 cereal adverts that only featured during the children’s hour
used creative strategies such as “animation and special effects as well as
‘bottom’ humour designed to appeal to young children”, whereas cereal products
which made nutritional claims dominated in prime time. The study further states
that some cereals made misleading claims concerning ‘fibre’, or claims of
‘energy’, ‘natural cane sugar’ and ‘wholesome goodness’. The research also
noted only eight diet-related PSAs (fewer than 0.2 per hour).
A time-trend analysis was also undertaken by comparing data from similar
studies from 1984 and 1986 (see Morton 1984). Although these trends reflect a
drop in advertising in the children’s hour (79.4 to 39.15 to 87.3), the number of
food adverts is increasing (from 38 to 44 to 130) which actually means that the
rate of food adverts per hour is also increasing (from 2.5 to 2.9 to 8.6).
Muehling
Kolbe, 1998
&
USA
Content analysis
582 prime-time adverts and
552 children’s adverts
A-51
The content analysis of television advertising found that disclaimers in children’s
advertisements featured most often in toy adverts (38%), breakfasts (22%) and
fast-food restaurants (17%), compared with top prime-time disclaimers for food
& snacks (22%), medicines (16%), and automobiles (15%). The study also
reported that informational disclaimers (that explain what the product does do)
were more likely to feature in adverts for toys (65 from 130), breakfasts (49 from
76), candy & gum (13 from 18), but that restrictive disclosures (that explain what
the product does not do) were more likely to feature in adverts for fast-foods
(61/59) (by these four categories – informational 65/130, 49/76, 13/18 and 30/59
– restrictive 39/130, 11/76, 0/18 and 61/59). No statistics provided as
percentages as adverts may contain more than one type of disclaimer.
From the data reported in this study, it is possible to extract content analysis
data of children’s advertising by product category: toys, 130; breakfasts, 76; fastfood restaurants, 59; snack food, 24; candy & gum, 18; and others, 38.
Comparisons can be drawn with content analysis data of prime-time advertising
by product category: food & snacks, 88; medicines, 66; automobiles, 62;
restaurants/retailers, 50; personal & beauty, 43; and others, 93.
Neville, Thomas
& Bauman, 2005
Australia
Content analysis
390 hours of television
broadcast during children’s
viewing times and 346 hours
of confectionary and fast-food
advertising
during
adult
programming.
Half of all food adverts were for products high in fat and/or sugar. Confectionary
and fast-food were the most advertised food categories during children’s
television hours. Confectionary adverts were three times more likely and fastfood adverts twice likely to be broadcast during children’s compared with adult’s
programming.
Ogletree et al.,
1990
USA
Content analysis
675 adverts yielded from nine
Saturday
mornings
of
children’s television
The content analysis of advertisements by product category found that 60.6% of
advertisements were for food products. In terms of the gender characteristics of
these food advertisements, most of the adverts had male narrators, consumers
or characters.
Although only 22 advertisements were rated as “enhancing the appearance of a
person, doll or animal”, more female than male main and supporting characters
were found in these advertisements.
Outley
&
Taddese, 2006
USA
Cross-sectional
content analysis
Three sets of television
adverts from 3
stations showing children’s
television programming (Black
Entertainment Television, The
WB
[Warner Bros], and Disney
Channel) during a 1-week
period, 11-15 July 2005, from
1500 to 2100 hours.
256 of 1098 recorded advertisements were food and beverage commercials.
Results indicate that 36.3% of all commercials were based on fast food
restaurants, 31.3% were for drinks, 16.8% were for candy, 13.7% were for
cereals, and 2.0% were for snacks (percentages do not total 100 because of
rounding). Compared with The WB and Disney Channel, Black Entertainment
Television had significantly (P=.001) more food and beverage advertisements.
Few health-related content (HRC) or physical activity–related content (PARC)
adverts were shown. Of 256 food and beverage commercials, only 8.2%
contained HRC and 9.4% had PARC. The HRC and PARC scenes contained
messages that were implied vs explicitly talking about the health or physical
benefits of the product.
Page
&
Brewster, 2007a
USA
Cross-sectional
content analysis
147 distinct food commercials
from 59 hours of recorded
children’s
television
broadcasting on US terrestrial
broadcast networks recorded
on 15 separate Saturday
mornings made up of 3 blocks
Findings show that the most frequently used promotional strategies were the use
of jingles/slogans (64%), showing children with the food (54%), and the use of
product identification characters (44%). The use of animation (63%), “real
children” (66%) and animated main characters (57%) were the most used
attention elements in the commercials.
A-52
from each of 5 networks
between 18 September 2004
and 19 February 2005. (Same
sample as Page & Brewster
2007b.)
Page
&
Brewster, 2007b
USA
Cross-sectional
content analysis
147 distinct food commercials
from 59 hours of recorded
children’s
television
broadcasting on US terrestrial
broadcast networks recorded
on 15 separate Saturday
mornings made up of 3 blocks
from each of 5 networks
between 18 September 2004
and 19 February 2005. (Same
sample as Page & Brewster
2007a.)
The most prominent emotional appeals were fun/happiness (85%) and play
(59%) followed by fantasy/imagination (50%), social enhancement/peer
acceptance (34%), and coolness/hipness (27%). Many of the products used the
term ‘super-charged’ (24%) or a similar adjective to describe the powerful taste
or other physical properties of the product. More than one-third of all the
commercials used a fruit appeal or association (35%). Statements or depictions
that a product was healthy or nutritious were quite rare among the commercials.
Powell,
Szczypka
Chaloupka,
2007a
USA
Cross-sectional
weighted
content
analysis
224,083 national ads from
170 top-rated TV shows
viewed by children aged 2
through 11 years from 1st
st
September 2003 to 31 May
2004.
In 2003-04, 27.2% and 36.4% of children’s exposure to total nonprogram
content time and product advertising, respectively, was for food-related products
(cereal, snacks, sweets, beverages, fast food restaurants, non–fast food
restaurants, and other food products). Similar distributions were found by race.
Cereal was the most frequently seen food product, making up 27.6% of all food
ads. Comparisons with previous studies suggest that, over time, food ads
account for a smaller share of the product ads seen by US children.
USA
Cross-sectional
content analysis
Recorded sample of top-rated
television
shows
(using
ratings data) to examine the
nutritional content for fat,
saturated fat, sugar, sodium,
and fiber of food-product
advertisements
seen
on
television by both children
and adolescents.
For 2-11-year-olds and 12-17-year-olds, respectively, a sample of 50351 and
47955 30-second-equivalent food-product advertisements and their related
nutritional content were weighted by television ratings data to provide actual
exposure measures of the nutritional content of food advertising seen by children
and adolescents.
Powell
2007
et
&
al.,
97.8% and 89.4% of food-product advertisements viewed by children 2 to 11
years old and adolescents 12 to 17 years old, respectively, were high in fat,
sugar, or sodium. On average, 46.1% and 49.1% of total calories among the
products advertised came from sugar in the advertisements seen by these
respective age groups. A total of 97.6% of cereal advertisements seen by
children 2 to 11 years old were for high-sugar cereals.
No substantial differences were found in the nutritional content of advertisements
seen by black and white children 2 to 11 years old. However, a slightly higher
proportion of food advertisements in general and across all food-product
A-53
categories seen by black versus white adolescents were for high-sugar products.
Powell,
Szczypka
Chaloupka,
2007b
USA
&
Rajecki,
McTavish
&
Jessup, 1994
USA
Cross-sectional
content analysis
Content analysis
Television ratings were used
to examine the distribution of
food advertising exposure
among adolescents aged 1217 years based on 170 toprated shows across network,
cable and syndicated TV
stations over the 9-month
period from September 2003
to May 2004.
A total of 238,353 30-second equivalent advertisements on the top-rated shows
were assessed.
114 adverts from 45 hours of
major television networks, 30
hours of cable and 25 hours
of local
The adverts were categorized by meal-type: Dry cereals (n = 88); toasted
products (n = 7); canned pasta (n = 6); additives (n = 5); hot cereals (n = 5);
frozen dinners (n = 2); and one soup advert. Of these 96.5% made reference to
breakfast or another meal-time. Twenty-two adverts did not contain stories,
leaving a final sample of 92 adverts for story-line analysis.
The results showed that among total nonprogram content time, food-related
products accounted for roughly one fifth (19.6%) of advertising exposure.
Excluding TV promotions and public service announcements, as a proportion of
all product advertising, total food-related advertising made up 26% of advertised
products viewed by adolescents. By race, the proportion of advertising
exposure to food products was 14% greater for African-American versus white
adolescents and total exposure to food advertising would be even larger for
African-American teens given that, on average, they watched more TV. Fast
food was the most frequently viewed food product category comprising 23% of
all food-related advertisements among adolescents.
When examining creative strategies, the study identified seven story themes:
• achievement
• conflict
• dependence – significantly more live action characters
• enablement
• mood alteration
• trickery
• violence – significantly more animated characters.
These seven varied by creative strategy and produced six clusters of themes,
with 64% of adverts clustered on some combination of violence, conflict and
trickery. The study also found four subtext themes:
• traditional
• practical
• emotional
• analytical.
All six story clusters were found to have a strong emphasis on emotional
subtext, less on traditional and practical and lacked analytical themes.
A-54
The study provides an example of a high emotional/low analytical advert for
Kellogg’s where a boy subverts his parents by playing a violent video game as
they comment on his sensible eating. One new story-line in an advert is
generated for each hour of children’s viewing. Concludes that the negative
themes of social strife, dishonesty and substance use are used in advertising of
this kind.
Reece, Rifon &
Rodriguez, 1999
Roberts
&
Pettigrew, 2007
Robinson, 2008
USA
Australia
Asia Pacific
(Fiji; China,
Hong Kong
SAR; India;
Indonesia;
Malaysia;
Nepal;
Philippines;
and Thailand)
Content analysis
Cross-sectional
content
analysis
using
quantitative
and
qualitative
methods
Cross-sectional
content analysis
The study used audience
figures to determine when
children watch the most
television and listings to
identify children’s
programmes: essentially
Saturday morning television
and cable channels.
Recordings were made during
February and March 1997
(different channels were
recorded each week in order
to get more adverts and
PSAs). This yielded 908
adverts, of which 416 were for
food.
The content analysis provided a breakdown by food product category: cereal,
39.2%; fast-food restaurants, 21.6%; candy/snacks, 21.4%; beverages, 4.8%;
other breakfast, 4.6%; other, 4.3%; pasta, 2.4%; and PSAs, 1.9%.
An analysis of the nature of food advertising found few celebrities featured in
advertising, but instead there were many cartoon characters. Cartoon images (or
partial animation) were a feature of nearly half of the advertisements. Taste (n =
40) and fun (n = 37) were the main theme appeals used in food advertising.
Time trend comparisons are made with Gussow (1972), Brown (1977), Cotugna
(1988) and Kotz & Story (1994). The study reports rising trends in the promotion
of foods, however this is difficult to assess because of network changes. Of all
adverts 45.8% were for food products; significantly lower than previous studies.
The study also reports a rise in advertising for fast food.
A medium scoring study of both the extent and nature of food promotion to
children.
28.5 hours of children's
morning
television
programming over 3 weeks in
Perth. 950 advertisements
were aired, 212 of which were
for food products.
Food advertising comprised 22.3% of ads sampled and 30 food campaigns were
identified. The most frequently advertised foods were those cautioned in the
Australian healthy eating guidelines (NHMRC 2003), with advertising messages
communicating numerous themes disregarding healthy eating practices.
Examination of marketing of
unhealthy food products (by
TV advertising, packaging,
innovative
marketing
techniques) by 6 multinational
food and drink companies
(choice based on market
share) between April and
June 2008. Report compiled
Multinationals were chosen based on their market share and foods/drinks
classified as unhealthy using UK Foods Standards Agency Criteria traffic
labelling scheme (high score in at least one of sugar, fat or salt content).
A-55
The qualitative themes evident in the ads were the prevalence of grazing, the
denigration of core foods, exaggerated health claims, and the implied ability of
certain foods to enhance popularity, performance and mood.
The techniques that these six multinational companies (Coca-Cola, Kellogg’s,
KFC, McDonald’s, Nestlé and PepsiCo) used to market their unhealthy products
to children in Asia Pacific countries included the following:
sponsorship of children’s sporting activities in schools, television advertising,
interactive websites, sports-themed contests and competitions, discounts and
from Consumers International
partner
countries’
data
(methodology
based
on
Which? 2008)
money-off vouchers for other products, use of cartoon characters, on-pack
games and promotions, movie tie-ins, children’s clubs, free toys, children’s
meals, celebrity endorsements, children’s parties, posters, language
teaching, sponsorship of children’s TV programmes.
Rodd & Patel,
2005
United
Kingdom
Content analysis
Television adverts
On average, 24 adverts were shown per broadcast hour (accounting for 15.8%
of the total schedule time). Nearly 40% (34.8%) of adverts related to food and
drinks products and 95.3% of these were for products that were deemed to be
potentially “erosive”. The most frequently promoted food and drink products
were breakfast cereals with added sugar (26.3%), confectionary (23.7%), and
non-carbonated soft drinks (18.1%).
Schwartz et al.,
2008
USA
Cross-sectional
study
All
ready-to-eat
(RTE)
breakfast cereals (N=161)
available from 4 leading
cereal
manufacturers
between
January
and
February 2006 – using local
grocery stores and the
companies’
websites
to
identify all products.
Of the 161 cereals identified, 46% were classified as being marketed to children
(eg, packaging contained a licensed character or contained an activity directed
at children).
Multivariate analyses of variance were used to compare children's cereals and
nonchildren's cereals with respect to their nutritional content, focusing on
nutrients required to be reported on the Nutrition Facts panel (including energy).
Compared to nonchildren's cereals, children's cereals were denser in energy,
sugar, and sodium, but were less dense in fibre and protein. The proportion of
children's and nonchildren's cereals that did and did not meet national nutritional
2
guidelines for foods served in schools were compared using χ analysis. The
majority of children's cereals (66%) failed to meet national nutrition standards,
particularly with respect to sugar content.
t tests were used to compare the nutritional quality of children's cereals with
nutrient-content claims and health claims to those without such claims. Although
the specific claims were generally justified by the nutritional content of the
product, there were few differences with respect to the overall nutrition profile.
Overall, there were important differences in nutritional quality between children's
cereals and nonchildren's cereals.
Sjölin, 2008
Sweden
Review
European data
Studies have documented that a high percentage of adverts targeting children
feature sweets, fast foods, and snacks and that exposure to such advertising
increases consumption of these products. The problem is that the great majority
of foods that are advertised are high in fat, salt and/or sugar (CFC, 2007).
A survey, made by Consumers International 1996, measured the amount of ads
directed to children in Australia, US and eleven European countries. The
countries that had the most TV adverts were Australia, US and UK. Sweden and
Norway had the lowest amount (Ekström & Sandberg, 2007). The OECD
countries that had the most TV ads 1996 were in descending order: Australia,
A-56
United States, United Kingdom, France, Greece, Finland, Germany, Denmark,
Netherlands, Belgium, Norway (TV 3*), Austria, Sweden (TV 4*). The average
number of food adverts per hour ranged from 1 to 12 (CFAC, 2007) (* No
adverts in other channels).
80% of all food adverts, when children watch terrestrial channels, are for
HFSS12. A Swedish study has shown that 80% of the food adverts are shown
between 1800 and 2100 hours during documental soap operas (28%), TV-series
(26%) and children’s programs (24%). The results from the study on TV ads
were similar to the internet study showing that no ads were found for fruits and
vegetables, meat, fish poultry egg and sausages. There were no soft drink
adverts and no adverts for ice cream and sweet desserts, cookies and bakery
goods, which might be a result of self-regulation. Of the food TV ads 13% were
for sweets, chocolate and crisps and 11% for fast food chains (Ekström &
Sandberg, 2007).
During children’s television hours, confectionary and fast foods adverts are most
frequent during all timeslots and especially children’s weekend viewing morning.
Solomon et al.,
1982
USA
Content analysis
Study 1: 37 adverts were
recorded from three 1-hour
segments of children’s
television on three
consecutive Saturday
mornings in November 1978
Study 1: Of a total of 37 adverts, 63% were for toys, and 35% for food.
Study 2: Of a total of 130 adverts in 1978, 86% were toys and 6% were for food;
in 1979, 71% of advertisements were for toys and 21% for food.
The findings from these studies demonstrate that food is advertised less (and
toys advertised more) outside the main television networks.
Study 2: 130 adverts were
recorded from six 1-hour
segments
of
weekday
children’s
programming
during September 1978 and
November 1979
Stern & Harmon,
1984
USA
Content analysis
Children’s
programming
recorded over 6 weeks
yielded 976 advertisements
The study provides a breakdown of commercial advertisements by product
category (the figures in brackets represent the percentage of advertisements
featuring a disclaimer):
Breakfast cereals, n = 339 (88.5%); confectionary, n = 114 (0%); toys, n = 105
(58.1%); restaurants, n = 50 (10.7%); health care, n = 48 (0%); clothing, n = 9
(0%); soft drinks, n = 3 (0%); other foods, n = 34 (0%); and other products, n =
106 (0%). The study finds that breakfast cereal advertisements are most likely to
have disclaimers of the “part of a nutritious breakfast” kind. Apart from
restaurants, disclaimers were absent from advertising for all other foods.
A-57
Taras & Gage,
1995;
Taras et al.,
2000
USA
Content analysis
95
hours
of
children’s
programming yielded 845
advertisements
Study 1 (Taras & Gage, 1995): A content analysis of advertisements from the
60 hours of weekday programming and 35 hours Saturday morning television by
product category is reported (note: 6 hours of this programming was excluded
from analysis as it was not considered child-oriented e.g. news and sports).
During these times there were 21.3 advertisements per hour (mean 28.6
seconds). Of these 47.8% were for food products, with 91% for foods high in fat
sugar or salt. Only 2.5 minutes (approximately 0.2% of all non-programming
time) contain PSAs with a nutritional message.
The data were compared with earlier studies (Barcus & Wolkin, 1977; and
the1980s data sweep study of Condry, Bence & Schiebe, 1988) that were
conducted prior to imposed regulations that limited advertising time to children.
The time trend analysis found that the percentage of adverts for cereals and
sweet snacks has decreased (the latter marginally), but that advertising for dairy
products had increased. Therefore, the amount of advertisements for products
high in fat, sugar and salt remained unchanged (e.g. comparing data from the
present study with the 1978 data, the percentage of adverts featuring products
high in sugar was 69.9% and 68.0% respectively, those featuring products high
in fat was 39.7% and 35.0% respectively, those featuring products high in salt
was 20.4% and 17.0% respectively, and finally those featuring products low in
salt was 8.9% and 11.0% respectively). The study concludes that during 1993
children watched more commercials of shorter duration, but that the message
content of the advertising was largely unchanged.
Study 2 (Taras et al., 2000): This study provides a more detailed breakdown of
advertising by food product category (17 food groups): high sugar cereals (n =
222); restaurants (n = 153); chocolate (n = 85); low-sugared cereal (n = 71);
fruit juice (n = 68); candy (n = 40); dairy (n = 36); sugared milk (n = 34); cookies
(n = 33); jelly (n = 24); soda (n = 23); cakes (n = 13); gum (n = 13); cheese (n
= 6); other beverages (n = 6); and soup (n = 1).
The study also compares this breakdown with parental beliefs about the effects
that such advertising has on children’s requests for food products (and parent’s
compliance with such requests). The study reports strong correlations between
the extent of food advertising and parental beliefs, concluding that certain
classes of food are requested by children and purchased by parents in the same
relative frequency as they are advertised.
Temple, Steyn &
Nadomane,
2008
South Africa
Cross-sectional
content analysis
Children’s programs were
recorded on two TV channels:
1. SABC2 was recorded from
0900 to 1100 h on six
occasions. The languages
A-58
There were 408 advertisements, of which 69 (16.9%) were for food. Two
categories:
A: Thirty-eight advertisements (55%) were for foods of generally poor nutritional
value: fast-food restaurants (n= 18; 12 advertisements were for McDonald’s and
used are mainly English and
Afrikaans plus a small amount
of Xhosa.
2. YoTV on SABC1 (1500 to
1630 h) is mainly in English
plus a small amount in Zulu.
Recorded during weekdays
between June and October
2006.
5 for KFC), highly refined breakfast cereals (n= 9), sweets (candies, n=6), potato
chips (n= 2), and sugar-rich cold drinks (n=2).
B: Twenty-nine advertisements (42%) were for foods of generally good
nutritional value, namely yogurt (n=24) and peanut butter (n=5), i.e., these foods
are nutritionally superior to foods in group A but, depending on the formulation,
are still likely to be inferior to comparable whole foods such as milk and nuts
Findings reveal that many advertisements were for foods of generally good
nutritional value, namely yogurt (24 advertisements) and peanut butter (5
advertisements); these two foods comprised 42% of all advertisements for food.
Nevertheless, 55% of food advertisements were for foods of poor nutritional
value. There were no advertisements for fruit, vegetables, or whole grains.
Our findings must be viewed with caution. We recorded children’s programs on
two TV channels, but only one of them had food advertisements. Our data are
based on 69 food advertisements, all from one TV program shown on weekdays
over a 5-mo period.
Thompson et al.,
2008
USA
Cross-sectional
content analysis
Food
and
drink
advertisements aired during
after-school hours (1500-2100
hours) on 2 US Spanishlanguage television stations
were sampled over a 1-week
period in the spring of 2006.
Reviewed 60 hours of programming. Of the non-program content, 47% was for
product advertisements, 15% (n = 153) of which was for food/drink. A mean of
2.5 food/drink commercials aired per hour (range 0-8), and the median duration
was 30 seconds; 31% of food/drink commercials advertised fast food, and 27%
advertised drinks, most (54%) of which were sugared. About one third (31%) of
the food/drink commercials targeted children, 12% featured Latino celebrities,
and 19% made reference to Latino culture Only 16% of the food/drink
commercials had health-related content.
Warren
2007
USA
Longitudinal content
analysis
Programming from November
2004 to February 2005 from
1400 to 2200 hours was
recorded and pooled with a
second
sample
of
programming from January to
May
2006.
1,880
30s
food/beverage adverts in first
sample and 2,898 in the
second.
In both Year 1 and Year 2, the most frequently advertised products are the
same: pizza/fast food (Yr1 34%, Yr2 27%); sweets (Yr1 12%; Yr2 15%); family
restaurants (Yr1 11%, Yr2 10%); breakfast foods (Yr1 10%; Yr2 10%); and
convenience meals/entrees (Yr1 6%; Yr2 8%). The results suggest few changes
in food advertising seen by children.
Programming
from
6
broadcast and 5 cable
networks recorded from 1400
Sample yielded 4,324 commercials: 51% child-targeted and 49% targeted to
general audience.
Warren
2008
et
et
al.,
al.,
USA
Cross-sectional
content analysis
A-59
When limited to child-targeted adverts, most of the same persuasive appeals are
most frequently used: taste/flavour (Yr1 26%; Yr2 31%) and mood alteration
(Yr1 18%; Yr2 15%). New appeals that appeared in the first year (Yr1 8%) were
replaced by nutritional (Yr2 7%) and adventure (Yr2 7%) appeals in the second
year (χ2=115.82, DF=18, p<.001).
to 2200 hours over 77 random
days from January to May
2006.
(same 2006 sample
Warren et al 2007)
as
Unlike previous content analyses, the results show that nutrition appeals are
among the most frequently used. However, as in past research, unhealthy foods
are most frequently advertised. 74% of all food/beverage products marketed are
among the 5 most unhealthy food categories: pizza/fast food (24%), sweets
(16%), breakfast-foods (13%), family restaurants (12%) and convenience meals
2
(9%) (χ =358.09, DF=12, p<.001).
Further, child-targeted commercials employ production techniques
persuasive appeals that children have found difficult to evaluate critically.
Weber, Story &
Harnack, 2006
USA
Cross-sectional
content analysis
Which?, 2008
United
Kingdom
Exploratory review
The top five brands in eight
food
and
beverage
categories, 40 brands in total,
were selected based on
annual sales data from
Brandweek
magazine’s
annual “Superbrands” report.
Examination of leading food
and
drink
companies’
marketing foods high in fat,
sugar and/or salt targeted at
children (under 16 years) via
websites,
packaging,
advertising
and
other
promotions from January to
June 2008. BARB data for
children aged four to 15 and
Nielsen data were analysed
for the period: 31 March–13
April 2008.
A-60
and
The results show a wide variety of Internet marketing techniques and advertising
strategies targeting children and adolescents. “Advergaming” (games in which
the advertised product is part of the game) was present on 63% of the Web
sites. Half or more of the Web sites used cartoon characters (50%) or
spokescharacters (55%), or had a specially designated children’s area (58%)
with a direct link from the homepage.
Since the last analysis (Which? 2006), many companies were still reliant on
traditional marketing techniques such as competitions and cartoon characters on
packaging. Others have shifted their focus to the teen market, using new media
such as websites (e.g. social networking) and mobiles, in a way that parents are
unlikely to be aware of. Other techniques include: online offers, viral marketing,
mobile marketing (e.g. ringtones), licensed characters on packaging and
websites, film tie-ins, sports schemes and sponsorship and misleading labelling
and disclaimers.
Since 2006: there has been a reduction in the use of licensed characters by the
leading companies in the UK.
TV advertising: Since 2006, stricter TV advertising regulations came into force in
the UK (manufacturers of foods high in fat, sugar and/or salt can no longer be
advertised during children’s programmes) but these are failing to stop children
being exposed to less healthy food advertising as family programmes are not
affected by the new TV advertising restrictions: during the 2 week period, 16 of
the top 20 programmes showing adverts that were watched by the highest
numbers of children were not affected by the restrictions.
Wilson, Quigley
&
Mansoor,
1999
New Zealand
Content analysis
42
hours
of
children’s
programmes
(Saturday
morning television: 08.00–
11.00 and the children’s hour
(15.30–18.30) yielded 269
food adverts (29% of total
adverts)
42 hours of children’s programmes, yielded 269 food adverts (29% of total
adverts) of which 63% were for products high in fat and/or sugar. The next most
frequently advertised products were fast-food items at 14%. A total of 76% of
food adverts were for food not eaten as part of a meal. There were no adverts
for low-cost healthy foods or traditional healthy Maori foods.
The study concludes that, even allowing for other food intake, a child who
consumed the advertised diet would have too much (based on recommend daily
intakes, RDI) fat and other energy foods (sucrose, fructose, and glucose) and
too little of a range of minerals and vitamins. The study also raises the
implications for the food security of poor people and ethnic groups.
Wilson et al.,
2006a; 2006b
New Zealand
Content analysis
858 adverts taken from 155
hours
of
children’s
programming
Compared to Australian channels, both New Zealand channels had significantly
higher proportions of food adverts that were classified as being high in fat and or
sugar (54% compared with 80% and 69% respectively). Of the food adverts on
New Zealand television 70.3% were for foods “counter to improved nutrition”
(95% CI, 67.1%, 73.3%) compared with those favouring improved nutrition at
5.1% (95% CI, 3.8%, 6.9%). The number of food adverts per hour was higher in
2005 than 1997 for the channel for which there was time trend data.
Winick
1973
USA
Content analysis
236 children’s adverts
A content analysis of the adverts obtained provides the following breakdown by
product category: cereals, 19.9%; candy, 15.7%; drinks, 15.7%; food for meals,
8.5%; restaurants, 6.8%; cookies, 6.4%; cakes, 5.9%; puddings, 5.5%; vitamins,
4.7%; snacks, 4.7%; non-edibles, 6.4%. A content analysis undertaken by
observing 92 of the 236 adverts identified on weekday television (between 07.00
and 14.00 h) across four weekends between September/October 1971 provides
the following breakdown of advertising by product category: cereals, 26.1%;
candy, 18.5%; drinks, 13.0%; food for meals, 9.8%; restaurants, 8.7%; cookies,
8.7%; cakes, 2.2%; puddings, 3.3%; vitamins, 5.4%; snacks, 3.3%; non-edibles,
1.1%. These observations weighted by exposure: cereals, 26.6%; candy, 12.7%;
drinks, 17.1%; food for meals, 7.0%; restaurants, 2.9%; cookies, 7.2%; cakes,
4.4%; puddings, 5.5%; vitamins, 13.2%; snacks, 2.8%; others 0.7%.
et
al.,
In terms of creative strategies, the study addressed 143 dimensions collapsed
into seven indices:
1. Product information – i.e. disclosure/disclaimer.
This scored second highest overall, most often for candy and the main candy
appeal.
2. Fantasy – animation, puppets, etc.
This scored third highest overall, most often for restaurant – fast-food
A-61
experience.
3. Personalities – use of real celebrities or cartoon characters.
Scored lowest overall, again most often in restaurants – no need to mention
ingredients.
4. Sales techniques – e.g. fun jingles, testimonials, "pester-power".
This was in the middle overall, yet again highest for restaurants.
5. Self social status – enablement, popularity.
Scored third lowest overall, highest for vitamins (snacks amongst foods).
6. Nutrition/health – good eating claims.
Scored second lowest, again highest for vitamins (though cereals by far the
most of food).
Note: was zero for candy, cookies (taste) and restaurants (eating experience).
7. Realistic social perspective – e.g. family settings.
Scored highest overall and in all categories except candy.
The factors analysis identified 36 factors; the first was animated/cartoon/fantasy,
and the second health/nutrition.
Young
1996
et
al.,
United
Kingdom
Review article
International
literature
academic
This is the best review identified in the current systematic review (refers to
Hastings et al 2003). The included studies are examined in detail, which
included making assessments of their methodologies (such as sample sizes,
and representativeness) and the findings. The review highlights the absence of
standardized procedures, such as data collection, coding, rating and analysis,
which make comparisons and trends difficult to assess.
It also looks for gaps in the research, such as the excess of studies conducted in
the 1970s focussing on sugar rather than other food constituents.
Young,
1990
1987;
United
Kingdom
Content analysis
Saturday morning television
and the children’s hour were
recorded over 47 days in
1983 and 1984 yielding 1750
adverts
The content analysis yielded 1750 adverts of which 573 (33%) were for food. Of
these, the study notes that ‘only’ 33% were sugared (N.B. products were defined
as sugared if over 10%). The author notes that this is much less than figures
from the USA at this time which range from 59–76% sugared foods. The preChristmas effect is also acknowledged, with toys featuring in over 10% of
adverts. Fifty-eight sugared product commercials aimed at children were
analysed (108 discarded if ‘obviously adult’ – no reference to children or no
child-centred techniques used). In this study, in cases where categorization of
adverts was difficult, resolution was only sought through discussion with another
colleague.
The following percentages describe the relative extent of different creative
strategies in the adverts examined: animation, 64%; consumption, 53%; family,
29%; puppets, 9%; day-life setting, 40% (fantasy 41%); normal action, 62%
(31% magic); humour, 57%; sugar, 9%. The following percentages describe the
A-62
relative extent of different creative appeals in the adverts examined: hyperbole,
26%; adventure [not stated]; fast-pace, 12%; romantic–domestic, 9%; romantic–
past, 16%; rhyme, 41%; metaphor, 29%; pun, 24%; and 21% use of a wellknown character.
A factor analysis was also conducted, although statistics are not reported, and
the analysis was undertaken on only 58 adverts, producing three vague factors:
1. Fast-paced child centred with extrinsic disclosures/disclaimers.
2. Extrinsic disclosures/disclaimers and rhyme.
3. Humour.
Zuppa, Morton &
Mehta, 2003
Australia
Content analysis
544 adverts from 63 hours of
programming classified as
specifically
produced
for
children 6–13 years of age or
suitable for children to view
without adult supervision
A-63
Of the adverts, 31% were for core foods and 79% were for non-core foods. Fastfoods, chocolate, and confectionary made up almost 50% of the food adverts
shown on television.
ANNEX 3
Data Extraction Table: The Effects of Food
Promotion to Children
Shading indicates a new study published between 1st March 2006 and 15th November 2008
STUDY
Aktaş
2006
Arnas,
COUNTRY
Turkey
STUDY DESIGN
Survey
SAMPLE
347
mothers
children
aged
years
with
3-8
KEY RESULTS
89.6% of the children either drank or ate something while watching
television and the food they consumed most while watching television
were fruits (60.8%), soft drinks (44.1%), popcorn/nuts (36.6%),
cake/pasta (33.7%), chips (33.4%) and candy/chocolate (28.8%).
CAPABLE OF
EST.
CAUSALITY?
No
Younger (preschool age) children paid more attention to advertisements
than the older children (i.e. 32% at age 5 and 1% at age 8). 32.85% of
children asked the parents to buy the products presented in the
advertisements while watching them, particularly 4-5 year olds (28% and
58% respectively and leavening off to 9% at age 6).
When out shopping, 40.3% of the children asked their parents to
purchase the goods that they saw on the television adverts and that
8.9% of them argued with their parents and/or cried in order for their
parents to buy that particular product. 3-6 year olds were more insistent
on crying for advertised products while out shopping however there was
no correlation with the time spent watching TV. Children tended to
request more sweetened products such as candy, ice-cream, biscuit,
cake or soft drinks.
Arluk et al., 2003
USA
Cross-sectional
survey
7–12-year-olds
Using BMI, 39.8% of children were obese. A significant relationship was
found between childhood obesity and computer usage, television
watching, total hours in sedentary behaviour and maternal BMI.
A-64
No
Atkin,
1978
1975b;
Atkin, 1975c
USA
Observational
study
516
family
‘dyads’
(mother plus child or
father plus child)
USA
Cross-sectional
survey.
775
4 –7
students
th
th
grade
In 66% of situations, the child initiated the interaction by demanding
(46%) or requesting (20%) a cereal. Of the children 9% explicitly
identified the free gift as the main reason for wanting a cereal, and
observers indicated that up to a quarter of children appeared to make
their decision at least partly on the basis of the free gift. Children
mentioned nutritional considerations as the main reason in only 1% of
interactions.
No
Two indices, cereal advertising exposure index and candy exposure
index, were created by multiplying Saturday morning television viewing
by self-reported frequency of viewing specific adverts for the two types
of products. Cereal advertising exposure correlated with reported
consumption of eight heavily advertised cereal brands (r = + .41). The
correlation remained strong when grade, sex, socioeconomic status and
school performance were controlled for (r = + .37). There was also a
correlation, albeit weaker, between cereal advertising exposure and
consumption of five lightly advertised brands (r = + .27), and a strong
correlation between consumption of heavily and lightly advertised
brands (r = + .58). In families with no reported rules restricting snacking,
the partial correlation between consumption of cereals and cereal
advertising exposure was + .49. There was a moderate correlation
between exposure to cereal commercials and child anger after a cereal
request denial (r = + .20).
Yes
Cereal advertising exposure was moderately correlated with frequency
of requesting cereal purchases (r = + .32). Twelve percent of
respondents with ‘light’ cereal advertising exposure asked their mothers
to buy cereals a lot compared with 27% of respondents with ‘heavy’
cereal advertising exposure (no significance values are quoted). There
was no correlation between cereal advertising exposure and beliefs
about the nutritional value of sugar, and only a very slight correlation
between cereal advertising exposure and number of tooth cavities (r = +
.09).
Multivariate analysis, using path analysis procedures, of the
relationships between key variables and requesting that parents buy
cereals found a relationship between cereal advertising exposure and
more frequent asking for cereals (r = + .27). Cereal advertising exposure
was linked to cereal consumption both directly (r = + .30) and indirectly,
through requests (r = + .27) which were then correlated with
consumption (r = + .26). The two exogenous demographic variables in
the model, grade, and socioeconomic status, were not significantly
related to cereal eating.
A-65
Confectionery advertising exposure was correlated with consumption of
three heavily advertised confectionery products (r = + .29); this dropped
slightly to r = + .25 when controlling for school grade, school
performance, sex, and socioeconomic status. Correlations between
confectionery advertising exposure and consumption of lightly
advertised confectionery products were equally strong, suggesting that
respondents who viewed more confectionery adverts on Saturday
morning television tended to eat all kinds of confectionery more
frequently than lighter viewers. A modest correlation (r = + .10) was
found between exposure and quantity of confectionery eaten per week.
No relationship was found between confectionery advertising exposure
and beliefs about the nutritional value of sugar or with number of dental
cavities in the past year.
A total advertising exposure index was created from measures of primetime, teen-oriented and Saturday morning viewing (this measure did not
include reported amount of attention paid to advertising). This was
correlated with more general measures of food consumption, including
asking parents to visit fast-food restaurants. A correlation of r = + .30
(dropping to r = + .28 when controlling for grade, sex, socioeconomic
status and school performance) was found between the total advertising
exposure index and consumption of five frequently advertised foods
(crisps, soda, hamburgers, chocolate drinks and cookies). Identical
correlations were found for consumption of less advertised foods
(pretzels, hot dogs, ice cream, and cake). The relationship between
advertising exposure and consumption was stronger for girls than boys
(r = + .33 vs. + .20) and for children with parental snacking restrictions
than for those without (r = + .31 vs. + .24). Exposure was modestly
correlated with frequency of asking to visit fast-food restaurants (r = +
.17).
Overall, the study indicated that children who reported watching more
Saturday morning television more often asked for cereals, expressed
anger when requests were denied, and ate cereals. More than twice as
many ‘heavy viewers’ of Saturday morning television as ‘light viewers’
reported making cereal purchase requests ‘a lot’ of the time. Exposure
has a direct effect on amount of consumption as well as an indirect
effect mediated by requesting cereal products. The strength of the direct
effect may explain why purchase requests to parents were not found to
be a stronger mediating variable.
Authors’ conclusions: Children most exposed to television tend to
A-66
consume more foods, both advertised and non-advertised, than children
less exposed to television. “Since it is unlikely that heavy eaters of
purely advertised foods are motivated to watch more television than
other people, it seems justified to conclude that this minor relationship is
evidence of a flow of causality from viewing to eating rather than the
reverse sequence”. But “a conservative reading of the data suggests
that the effect is not strong”.
Barnabè et al.,
2008
Italy
Review
Existing
data
and
expertise, particularly
relating to Europe.
A significant positive correlation was found between overweight
prevalence and the promotion of energy-dense foods, but there is no
evidence that it may be due only to this factor. On the contrary, the
negative, but weaker, correlation with the promotion of healthier foods
may suggest a potential benefit for healthy food marketing actions which
may actually counter obesity.
No
Children’s nutritional knowledge, food preferences, purchasing and
purchase related behaviour, consumption, and diet and health status are
influenced by advertising. Several studies indicate that there is a
favourable attitude towards advertised products.
Even exposure to advertisements as brief as 30 seconds long can
significantly influence the food preferences made by children as young
as two years old. The influence of food promotion on children’s
behaviour is independent of other factors and causes both brand
switching and category effects. Even though a link was found, there is
no evidence, indeed, to proof that advertising has a direct effect on
children’s diet, and on obesity. Focusing on the effects of television
advertising, a relation between adiposity and television advertising was
confirmed but the evidence is not sufficient to support a causal
relationship. The results indicate that there is evidence of a relationship
between children food marketing and children behaviour, but it is not
sufficiently clear whether there is a relation between overweight or
obesity in childhood and children food marketing.
Barry & Hansen,
1973
USA
Non-randomized
non-controlled
experiment
nd
60 2
grade pupils
from
two
Dallas
schools. Pupils were
from two classes of 30
pupils each. One class
was all white and one
was all black
There were no significant recall differences between racial groups on six
of the eight items. On the other two items, which assessed brand recall,
black children had significantly poorer recall than white children.
The three preference items displayed significant racial differences. For
all three items, black children expressed a strong preference for one
advert, cereal or character over another, whereas white children did not
have decided preferences (P < 0.01).
A-67
No
Batada
Borzekowski,
2008
&
USA
Clustered design
with children and
schools randomly
selected. In-depth
interviews including
open
ended
questions
58 children (grades 35) and their parents,
randomly selected from
public
schools
in
Baltimore City (urban
low income area).
Half of the participating children (N=29, 50%) could describe
unprompted and from memory 5 or more of the 10 commercials
associated with the presented cereal characters/logos. Logo and
character recognition was negatively associated with a child’s weight;
overweight children were less likely than their average weight and at-risk
for overweight counterparts to recognise more characters. Having more
than 3 working televisions in the home was associated with higher
recognition of cereal characters /logos.
No
Fewer than half of the elementary school children were able to
accurately describe the persuasive intent of television commercials.
Children who were overweight wee more likely that average weight
children and children who were at-risk for overweight to understand the
persuasive intent of television commercials (χ2 =10.9, p<0.05).
Bolton, 1983
USA
Cross-sectional
survey
Data were obtained for
262 children aged 2–
11 years from twoparent families with a
television set.
The
families
were
predominantly
white
and
of
higher
socioeconomic status,
and
children
were
distributed
evenly
across age and sex
categories.
The children were part
of a large household
survey conducted in
Ohio in 1977.
No
information is provided
on how the household
survey was generated
other
than
that
households
were
“contacted
through
community
organizations in Ohio”.
The structural coefficient estimates indicated that the most important
direct influence on children’s behaviour was parental behaviour. Child
descriptors (age, missed meals, snacking, food commercial exposure)
were consistently of secondary importance. The direct effects of
parental supervision were mixed, and smaller in magnitude than the
effects of child descriptors.
The study uses structural equation modelling to predict the relationship
between food commercial exposure and children’s behaviour. (For
details of the equations, see original paper).
In equation 1, children’s food commercial exposure increased
significantly with parental food commercial food exposure and
decreased with parental supervision of television viewing. However,
parental supervision explained only 3% of variance in children’s food
commercial exposure, while parental exposure explained 22% of
variance. In equation 2, parental snacking frequency explained 29% of
variance in children’s snacking frequency. However, parental diet
supervision did not have the significant negative effect on snacking
which was hypothesized. Age had only a small impact on snacking,
explaining only 4% of variance in snacking (with older age being
associated with less snacking). Missed meals were not significantly
related to snacking frequency. Children’s food commercial exposure had
a significant effect on snacking frequency, although small (explaining
only 2% of the variance). The impact of parents’ snacking frequency
was 15 times larger than the impact of advertising.
In equations three and four, an increase in parental calorific intake
A-68
Yes
significantly increased children’s calorific intake (explaining 9% of
variance), and an increase in parents’ nutrient efficiency significantly
increased children’s nutrient efficiency (explaining 8% of variance). The
direct effects of parental behaviour were smaller in these two equations
than in equations one and two because of the indirect effects of parental
behaviour through children’s food commercial exposure and snacking
frequency. Parents’ diet supervision did not significantly affect children’s
calorific intake or nutrient efficiency. Snacking significantly increased
children’s calorific intake, explaining 5% of its variance, and significantly
decreased their nutrient efficiency (explaining 6% of variance). Missed
meals caused a decrease in calorific intake, although snacking had a
more detrimental effect than missed meals. In these two equations,
snacking and missed meals together had as great an effect as parental
example, suggesting the importance of dietary habits in determining
children’s nutritional well-being.
Age-related effect were generally greater than parental behaviour,
children’s snacking or children’s missed meals, explaining 12% of the
variance in children’s calorific intake and 8% of the variance in children’s
nutrient efficiency. Older children consumed proportionally less calories
than nutrients, causing nutrient efficiency to rise.
Children’s food commercial exposure did not have a significant direct
effect on children’s calorific intake, but did significantly decrease
children’s nutrient efficiency, although explaining only 2% of the
variance. Food commercial exposure also had significant indirect effects
on children’s calorific intake and nutrient efficiency, increasing the
former and decreasing the latter. This indirect effect worked through the
effects of food commercial exposure on children’s snacking frequency,
which in turn increases their calorific intake and decreases their nutrient
efficiency. Again, these effect sizes are small (around 1% of the
variance). The combined direct and indirect effect of food commercial
exposure on children’s calorific intake and nutrient efficiency was at
most half the size of the direct impact of other predictor variables.
In equation 5, parental nutrient balance explained approximately 9% of
the variance in children’s nutrient balance, whereas parental diet
supervision did not have a significant effect. In this equation, neither
children’s food commercial exposure nor children’s snacking had any
significant effects.
Overall, the analyses suggested that children’s exposure to television
food advertising influenced their diet in three separate ways. Firstly, it
A-69
significantly increased the number of snacks consumed. According to
the structural equation model, an increase in food advertising exposure
by an additional 25 minutes per week (12 hours total viewing per week)
would have caused a child to consume one additional snack per week.
Secondly, this additional snack would have increased the child’s calorific
intake by approximately 1.4 % and decreased the child’s nutrient
efficiency by a similar amount, assuming that children typically snack on
low nutrient, high calorie foods.
Thirdly, children’s exposure to
television food advertising significantly decreased their nutrient
efficiency directly, in addition to the indirect effect through increased
snacking frequency. According to the model, an increase in food
advertising exposure by an additional 25 minutes per week would have
decreased the child’s nutrient efficiency by about 6%. Because in this
equation calorific intake was not affected, this implies that the child
consumes low nutrient, high calorie foods in place of foods with
equivalent calories but higher levels of nutrients. This was consistent
with the notion that children’s snack preferences are influenced by the
low nutrient, high calorie foods advertised on television. The influence of
parental behaviour was greater than that of television advertising
exposure. However, parental behaviour was also, according to this
model, an important influence on children’s television food advertising
exposure.
Borzekowski &
Robinson, 2001
Boynton-Jarrett
et al., 2003
USA
USA
Randomized
controlled
experiment
Randomized
controlled trial
46 children aged 2–6
years from Head Start
programme
in
California
There were no significant differences between children in the two
experimental conditions in demographic or media use characteristics.
548 ethnically diverse
students from public
For each additional hour of television viewed per day, fruit and
vegetable servings per day decreased (−0.16) after adjustment for other
Yes
Children who viewed the commercials tape were more likely to select
the advertised food in seven of the nine pairs of items. The remaining
two items were the breakfast cereal, where children in both groups were
equally likely to choose the advertised product, and the remote control
toy, where children in the treatment group were not more likely to select
the advertised product. The largest odds ratios were for the two food
items advertised twice on the tape [biggest difference between
experimental and control groups e.g. three times more likely]. There
were no significant differences between boys and girls in the rate at
which they selected advertised vs non-advertised food items. There
were also no significant differences in the proportion of responses for
advertised items in relation to amount of media technology in the home
(based on ‘low’ vs ‘high’ numbers of televisions, VCRs and videogame
players in the home).
A-70
Yes
schools
Bridges
&
Briesch, 2006
USA
Longitudinal survey
Households in a multistore panel in Colorado
covering a 118 week
period starting in the m
id-1990s.
variables. Baseline hours of television viewed per day was also
independently associated with change in fruit and vegetable servings.
Finds evidence that advertising targeted at children (but not other
promotional methods tested, which included price, POS) results in
changes to food purchase behaviour at both brand and product category
level.
Yes
There is moderate evidence of effect of advertising on childreninfluenced food category and brand purchases (mediated through nag
factor and demand elasticity of household).
Buijzen,
Schuurman
&
Bomhof, 2008
The
Netherlands
Cross-sectional
multi-method
(diary/survey)
study
234
urban
and
suburban households
near Amsterdam with
children aged 4-12
years in 8 elementary
schools in Spring 2006
Children’s exposure to food advertising was significantly related to their
consumption of advertised brands (β=.21) and energy-dense product
categories (β=.19). The relationship between advertising exposure and
overall food consumption was considerably stronger overall in lowerincome families, but nonsignificant (β=-.08). The impact of television
food advertising generalised to other brands within the same product
category as the advertised brand Family communication about
consumer matters moderated the advertising and consumption variables
which were reduced by high levels of consumption-related
communication within a family.
Yes
Cantor, 1981
USA
Randomized
controlled
experiment using a
2 x 2 independentmeasure factorial
design
37 children aged 3–9
years (25 boys and 12
girls).
Mean scores for proportion of sweet dessert choices differed between
the conditions during the pre-experiment period. Analysis of variance
was therefore conducted, and indicated that none of the pre-experiment
differences between conditions were significant. Analysis of variance on
the data for the week of exposure yielded an interaction effect
approaching significance (F(1,22) = 3.45, P < 0.07). None of the
associated means differed significantly from one another in subsequent
tests (Newman-Keuls method). In order to correct for individual
differences prior to exposure to the manipulation, an analysis of
covariance was performed, using the pre-exposure proportions as the
covariate and the proportions after exposure as the criterion. This
yielded no significant main effects, but an interaction effect of borderline
significance (F(1,32) = 3.68, P < 0.06).
Yes
Multiple t-tests (Tukey’s) indicated that children exposed to the serious
PSA plus the toy advert (i.e. not exposed to the sweet dessert advert)
made significantly fewer sweet dessert choices after the experiment
compared with children exposed to the serious PSA plus the sweet
dessert advert and children exposed to the humorous PSA and the toy
advert (P < 0.05). Data for proportion of fruit choices followed a similar
pattern. Children exposed to the serious PSA plus the toy advert made
A-71
significantly more fruit choices after the experiment compared with
children exposed to the serious PSA plus the sweet dessert advert and
children exposed to the humorous PSA and the toy advert.
The serious PSA appeared to have more impact on children’s fruit
choices than the humorous version when it was not followed by a sweet
dessert advert, although the lack of a main effect suggested that the
presence of the sweet advert did not in itself have an effect. However,
the group exposed to the combination of serious PSA and toy advert
made significantly fewer sweet dessert choices and significantly more
fruit choices, post-experiment, than the group exposed to the serious
PSA and confectionery advert.
Carruth
2000
et
al.,
USA
Longitudinal study
34 children 5-year-old
children (14 males, 20
females)
Carruth,
Goldberg
&
Skinner, 1991
USA
Self-completion
survey
887 10 –12
grade
students in four high
schools
th
th
The most commonly given reasons for product preferences were, in
descending order: liking/favourite, flavour/taste, characters or action
figures, product type, colour, the foods depicted on the package, prior
consumption, appearance, free gift, because parents buy it, and health
reasons. Some reasons increased over time and some decreased, but
no statistical analyses are reported. Responses to mothers’
questionnaire are not relevant to the review.
No
Overall, about 9% of respondents reported ‘very often’ or ‘often’ talking
about food adverts with their parents, and about 6% with their friends,
whereas 72% reported ‘never’ or ‘rarely’ discussing adverts with their
parents and 75% with their friends.
No
Eight percent reported that seeing a food advert made them want to get
something to eat ‘every day’, 14% ‘several times a week’, 27% ‘once or
twice a week’, 25% ‘less than once a week’, and 27% ‘never’.
Responses to an open-ended question about snacks were consumed in
front of the television indicated that crisps (55%), fizzy drinks (21%),
popcorn (21%), cookies (19%), sandwiches (18%), fruit (15%) and
sweets (14%) were most frequently reported.
The study reports that snacking in front of the television in this study
was not related to seeing a commercial, but no data are provided.
Carter, 2006
Australia
Systematic
literature review
Research, reviews and
commentaries
identified
through
Boolean searches of
PubMed, ScienceDirect
and Google; searches
Controlled studies have consistently shown that children exposed to
advertising choose advertised products at significantly higher rates than
those unexposed, and the effect does not diminish with age.
Two separate literature reviews have concluded that there is sufficient
evidence to suggest food advertising has an independent effect on the
A-72
No
of bibliographies of
relevant papers – with
preference given to
Australian data.
preferences, purchasing behaviour and consumption of food by children
at both brand and category level.
Multiple cross-sectional studies have demonstrated that children’s
television viewing is positively associated with higher intakes of dietary
energy through fat, sweet and salty snacks, and carbonated beverages
and thus are related, however studies fail to establish direction of
causation and co-variables, such as poor parental supervision, cannot
be ruled out as a causal mechanism.
From the Australian evidence, the review finds the link between
television and childhood obesity is weak, with a ~1% independent effect
size.
Chamberlain,
Wang
&
Robinson, 2006
USA
Prospective cohort
study
827
third
grade
children
from
12
elementary school in
northern
California
(Sep-Oct
1999);
a
ethnically and sociodemographically
diverse
prospective
sample of 386 pupils in
6
schools
were
followed up 3 times for
20 months (Apr-May
2000, Sep-Oct 2000
and Apr-May 2001).
Results relevant to the review (excluding the 6 schools in the smoking
prevention trial):
No
Children reported watching +10 hours TV and +22 hours total screen
time (TV, movies, videos or videogames) per week; and making nearly 1
request/week for toys seen on TV and more than 1 request/2 weeks for
food/drinks they had seen on TV. At baseline, children’s screen
exposure was significantly associated with requests for foods/drinks
(P<.001). In prospective analysis, children’s screen exposure at
baseline was significantly associated with their mean number of
foods/drinks requests 7, 12 and 20 months later (P<.01).
Overall, after adjusting for baseline requests and socio-demographic
variables, the relationship between screen media exposure and future
requests for advertised foods/drinks remained significant for total TV
viewing and total screen media exposure.
Chan & McNeal,
2003
China
Three
crosssectional surveys
1665
parents
of
elementary
school
children aged 6–14
years
in
Beiking,
Nanjing and Chengdu
Results indicate that Chinese parents hold negative attitudes towards
television advertising in general and children’s advertising specifically.
The negative attitudes result mainly from the perception that advertising
is deceptive and annoying. Parents feel strongly that advertising should
be banned on children’s programming. Ninety-eight per cent of parents
exercise some control over the contents and times of television viewing.
Parents perceived that they have a great influence on their children’s
attitudes towards advertising.
No
Chan, 2000
China
Semi-structured
interviews
448 Chinese children
aged 5–12 years in
China, Hong Kong
Results relevant to the review:
Favourite advert: 222 adverts were identified as favourites (not all
respondents could identify a favourite advert). The most frequently cited
No
A-73
SAR
(equally
distributed between the
sexes)
adverts were for food and drink (57%) followed by toys (12%), mobile
phones (8%), restaurants/fast-food (6%), and supermarkets (4%). All
other categories received 3% or fewer citations. The main reasons given
for liking adverts were categories as ‘entertainment’ (57%), liking the
product (23%), and choice of character/celebrity (18%).
Least favourite advert: 116 ads were identified as least favourites. The
top categories of least favourite adverts were mobile phones (18%),
food and drink (17%), PSAs and supermarkets (10% each), toys (7%)
and property (5%). All other categories received 4% or fewer citations.
Brand recognition was stronger for liked adverts than for disliked
adverts: 85% of respondents could spontaneously recall the brand
featured in their favourite advert, compared with 60% for their least
favourite advert.
Brand recall: when presented with three slogans (for Kinder chocolate,
Nestle milk powder, Maltesers), between 18 and 32% could recall the
brand. Older children had better brand recall than younger children.
Comprehension of the four adverts presented in storyboard form: 63%
of children identified the correct messages in both the Kinder chocolate
advert and the equal rights PSA. Less than half could identify the key
messages in the tea advert (40%) and taxi advert (27%). Understanding
of key messages was significantly related to school year, with increasing
understanding as children got older.
Chernin, 2008
USA
Single
factor
experimental study
Convenience sample
of 133 5-11 year olds
(60.2% female; M=8.18
years, SD=1.45) from
two elementary schools
in
suburban
Philadelphia.
Exposure to commercials (or a sweetened breakfast cereal and an
orange-flavoured drink mix was positively and significantly associated
with preference for the advertised product (β=.33, robust SE=.11, p≤.01)
when shown coloured pictures of the product and three others within the
same product category.
Yes
Age was not statistically significant (β=.07, robust SE=.07, p=.36);
younger and older children exhibited similar preferences for the product.
Boys were more influenced by the commercials than girls (β=.48, robust
SE=.22, p=.03). The gender difference may have been exacerbated by
the commercials’ main character being a boy.
The two commercials were equally persuasive as the interaction
between ad exposure and ad variable was not significant.
Clarke, 1984
USA
Randomized
controlled
experiment
80 preschool children
of both sexes with a
mean age of 53
months.
There were no significant differences found in any of the fruit drink
measures between groups depending on programme type, advertising
exposure, or whether or not food was offered during the viewing
situation. Significant effects were found in relation to the toy truck
A-74
Yes
questions, but these are not relevant to the review.
Coon et al., 2001
USA
Cross-sectional
survey
91 parent–child pairs
from Maryland, USA.
th
All children were in 4 –
th
6 grade (average age
10 years)
Televisions were more likely to be on during meals in households with
lower incomes (P ≤ 0.01), single parents (P ≤ 0.05) or less educated
mothers (P ≤ 0.05). Television presence during meals was inversely
related to parents’ nutritional knowledge, attitudes and norms (P ≤ 0.05)
and positively related to parents’ attachment to meat (P ≤ 0.01), and
frequency of parents preparing quick suppers (P ≤ 0.01). There was a
relationship between the television being on during two or more meals
per day and lower consumption by children of foods in the fruit and
vegetable group (fruit, vegetables, juice and juice drinks) (P ≤ 0.01).
Consumption of foods in the meat group (red meat, processed meat,
chicken, egg and fish) and foods in the pizza/salty snacks/soda group
was significantly higher among children exposed to television during two
or more meals per day (P ≤ 0.05 and P ≤ 0.01 respectively). Children in
this group derived 6% more (compared with children exposed to less or
no television during meals) of their daily total energy from all three meat
groups combined (P ≤ 0.01), 5% more from pizza, salty snacks and
sodas (P ≤ 0.01), and nearly 5% less from fruits, vegetables and juices
combined (P ≤ 0.001). They also consumed more caffeine than children
with low television exposure (P ≤ 0.01).
Multiple linear regression examined the relationship between each of
five dependent variables (children’s consumption of five food groups:
fruit; vegetables; red meat; pizza and snacks; soda) and the
independent variable presence of television during meals, controlling for
socioeconomic factors, parents’ nutritional knowledge, attitudes and
norms and parents’ use of quick foods. There was a significant
relationship between more exposure to television and higher
consumption of red meat (P ≤ 0.01), pizza and snacks (P ≤ 0.05) and
soda (P ≤ 0.05), and lower consumption of vegetables (P < 0.01). Of the
socioeconomic and demographic variables, only two were significant:
family income was significantly inversely related only to fruit
consumption (P ≤ 0.05), and being black was negatively associated with
soda consumption (P ≤ 0.05). Multiple regression analysis also showed
television during meals to be independently and significantly associated
with percentage total daily energy from the three combined food groups.
Compared to children with lower exposure to television during meals,
higher exposure children derived a lower percentage of total daily
energy from fruit, vegetables and juice (P ≤ 0.001) and a higher
percentage from meat (P ≤ 0.05) and from pizza, snacks and soda (P ≤
0.001). Of the socioeconomic and demographic variables, being black
A-75
Yes
was significantly associated with higher percentage of total daily energy
from fruit, vegetables and juice (P ≤ 0.01) and from meat (P ≤ 0.01), and
lower percentage from pizza, snacks and soda (P < 0.05). The study did
not report the proportion of variance in consumption explained by the
regression equations and it was unclear whether the reported regression
coefficients were standardized or unstandardized. This makes it
potentially difficult to assess the strength of influence of television during
meals relative to other influences. However, the regression analyses
showed that, when the socioeconomic and demographic factors were
controlled for, television during meals had an independent and
significant influence on frequency of consuming four foods (vegetables,
red meat, pizza and snacks, and soda). Furthermore, with the exception
of soda, none of the socioeconomic or demographic variables had a
significant influence. This shows that, for those particular foods,
television during meals had a stronger influence on consumption than
the socioeconomic and demographic variables. Similarly, with the
exception of race, television during meals had a stronger influence than
the other socioeconomic and demographic variables on the percentage
of total daily energy derived from the three combined food groups.
Dawson et al.,
1988
USA
Randomized
controlled
experiment
80 white, middle-class,
public
school,
kindergarten
(mean
age
6.2
years)
students residing in the
north-western USA
“All children (100%) engaged in some consummatory behaviours. Half
(50%) engaged in behaviours such as touching the food, picking up the
food, pretending to eat, or eating” (p.1356) [no further breakdown is
provided].
Children in the ‘low nutrition’ food stimulus condition
displayed more consummatory behaviours than children in the ‘pronutrition’ food stimulus condition (mean score 15.35 vs mean score
10.50, P < 0.01), regardless of the commercial shown. There was no
significant difference in consummatory behaviours by children exposed
to the different commercial types. Children exposed to the ‘low nutrition’
food commercial displayed 13.10 consummatory behaviours compared
with children exposed to the ‘pro-nutrition’ commercial who displayed
13.45 consummatory behaviours.
Yes
Self-reported temptation to transgress displayed a trend, with greatest
temptation being reported after viewing the ‘low nutrition’ commercial
followed by the ‘pro-nutrition’ commercial, the toy commercial and the no
commercial control. However, this was only significant at the 10% level
(P < 0.09). Girls reported significantly more temptation to transgress
than boys (mean score 3.41 vs mean score 2.08, P < 0.05).
Del
Toro
&
Greenberg, 1989
Puerto Rico
Cross-sectional
survey
th
th
225 9 –12 grade high
school students from
two schools (one public
Girls were significantly more positive and less negative about food
adverts than boys, and reported greater desire for foods seen in adverts
(P < 0.05 for all three factors). Girls and boys were equally likely to ask
A-76
No
and one private) in a
metropolitan area in
Puerto Rico.
One
class from each grade
level in each school
was surveyed.
The
sample contained twice
as many females as
males so data were
weighted to balance
sex across the grade
levels.
The sample
was described as 89%
Puerto Rican,
and
similar to the national
profile in terms of
number of people living
in each household
their parents to buy foods they had seen advertised (between 35% and
48%), and to buy advertised foods themselves. Younger respondents
(9th–10th grade) were more likely to ask their parents to buy foods they
had seen advertised than older children (P < 0.05). When reported
television behaviours were correlated with reported nutrition behaviours,
there was a consistent significant correlation between amount of
snacking and amount of television watched at different times (during the
week r = 0.21, at weekends r = 0.17, and during meals r = 0.19, P <
0.05).
Television viewing was consistently significantly related to holding the
opinion that eating is good for you.
There was a significant correlation between using one’s own allowance
to buy foods and television viewing on weekdays (r = 0.14), Sundays (r
= 0.22), and during the week (r = 0.25). There was a relationship
between watching television on Saturday morning and using one’s own
money to buy ‘bad’ foods such as chocolate confectionery, soft drinks
and potato crisps (r = 0.25, P < 0.05). Television exposure was not
related to the number of meals eaten or skipped, frequency of eating
out, intake from different food groups, opinions about what constitutes a
healthy diet or making purchase requests from parents.
Dickinson, 1997
United
Kingdom
Semi-structured
interviews
12 United Kingdom
households containing
at least one child aged
11–18 years and two
adults.
Limited findings reported on how households engage with food
advertising. Young people were described as “particularly adept at
recalling the voice-overs in food adverts almost verbatim”.
No
Dietz
&
Gortmaker, 1985
USA
Two
crosssectional surveys
and
one
longitudinal survey
Three samples were
included in the study:
Cross-sectional analysis of the cycle two sample (children aged 6–11
years) indicated a significant relationship between television watching
and obesity; children who watched more television experienced
significantly more obesity (P < 0.01) and super-obesity (P < 0.02) than
children who watched less television. There were no significant
relationships between obesity and children’s reported number of friends,
ability to get on with friends, time spent with friends, time spent alone,
listening to the radio, reading, or other leisure activities. Cross-sectional
analysis of the cycle three sample (children aged 12–17 years) also
indicated a significant relationship between television watching and
obesity; children who watched more television were significantly more
obese (P < 0.0001) or super-obese (P < 0.0001) than children who
watched less television.
Yes
6965 children aged 6–
11 years studied in
cycle two of the USA
National
Health
Examination
Survey
between 1963 and
1965
6671 children aged
12–17 years studied in
cycle three of the USA
A-77
National
Health
Examination
Survey
between 1966 and
1970
2153 children studied
in cycle three who had
also previously been
studied in cycle two.
There was a dose-response relationship between obesity, super-obesity
and time spent watching television. Estimated regression coefficients
indicated that the prevalence of obesity increased by 1.2 to 2.9% for
each additional hour of television watched per day. Similarly, the
prevalence of super-obesity increased by 1.4 to 1.6% for each additional
hour of television watched per day. When a range of control variables
were entered into the analysis to control for their potential influence on
obesity, super-obesity and television viewing – including past history of
obesity at cycle two and socioeconomic characteristics of the family, the
magnitude of the television–obesity relationship was not altered in the
cross-sectional analyses. Controlling for past obesity and
socioeconomic characteristics did reduce the influence of television
viewing on obesity in the longitudinal analysis, but the relationship
between television viewing and obesity and super-obesity was still
significant (P < 0.001 and P < 0.05 respectively).
A more stringent test of the relationship between television viewing and
obesity was obtained by examining the association between television
viewing in cycle two and obesity in cycle three (i.e. 3–4 years later), in
the longitudinal sample. When cycle two obesity and family
socioeconomic characteristics were controlled for, coefficient estimates
for cycle two television viewing and obesity and super-obesity in cycle
three were 0.008 (P < 0.07) and 0.006 (P < 0.03), i.e. marginally
significant.
Dixon
2007
et
al.,
Donkin et al.,
1992;
Donkin,
Neale & Tilston,
Australia
United
Kingdom
Cross-sectional
survey
and
experiment
Cross-sectional
survey
919 Grade 5 (mean
age 10.3 years,
sd 0.52 years) and
Grade 6 (mean age
11.3 years,
sd 0.5 years) students
recruited large
primary schools within
metropolitan
Melbourne
The index of weekly hours of TV viewing indicated children watched an
average of 17 h of TV per week (M=16.83, sd=8.93; range 0-31.5).
Controlling for gender, grade and SES, the survey showed that hours of
weekly TV viewing was significantly positively associated with more
positive attitudes toward fizzy drinks, chocolate and fast food, stronger
liking of junk food, perceiving that other children ate junk food more
often, perceiving junk food to be healthier, and higher reported
frequency of consumption of junk food.
507
parents
or
guardians of children
aged 7–11 years in the
Children’s food purchase requests reported by parents were for a range
of foods, the largest category being cereals (18%), followed by biscuits
and cakes (11%), fruit and vegetables (11%), sweets and chocolates
No
The experiment found exposure to the junk food ads did not enhance
children’s attitudes or intentions towards unhealthy foods. Exposure to
ads for nutritious foods promoted selected positive attitudes and beliefs
concerning these foods. This generally happened irrespective of
whether healthy food ads were aired alongside junk food ads or not.
A-78
No
1993
Donohue, 1975
USA
Semi-structured
interviews
Central
television
district of England
(10%), drinks (10%), and meat and meat products (9%). Eleven percent
of requests were specifically for Kellogg’s cereals. Forty-five percent of
the requested products had added sugar. A relationship between
television watching and sugar consumption by children is described in
the text, but it is not clear how this relationship was measured and
analysed.
162 black elementaryst
school children in 1 –
rd
3 grade
Favourite adverts were listed by product category. Food adverts were
the most popular, comprising 18% of favourite adverts for the sample as
a whole. Food adverts were listed as favourites by 9% of first grade
pupils, 32% of second grade pupils, and 15% of third grade pupils. They
were followed closely by programme trailers, toys and games adverts,
and adverts for medicines/vitamins. A fifth could not recall a favourite
advert, and 13% had no favourite. When asked what they specifically
liked about adverts, humour was the most important feature (27%),
followed by entertainment (18%), with smaller numbers giving ‘cartoons’,
‘information’ and ‘action’ as important features.
No
The majority of children (90%) reported that they helped their parents to
pick out items when shopping. When asked how they knew what to buy,
interpersonal influence (‘someone told me about it’) was the most
commonly given reason (34%), followed by having seen the product on
television (29%), having seen it in the store (13%), needing it (3%), and
other reasons (22%).
When asked whether the cereal itself or the free gift was more important
in selecting a cereal, both were equally important for boys, but for girls,
the free gift was the main consideration (56% free gift, 44% cereal).
First grade children appeared to put more emphasis on the free gift than
on the cereal. However, none of the differences were significant.
Escalante
de
Cruz et al., 2004
India;
Indonesia;
Malaysia;
Pakistan;
Philippines;
Republic
of
Korea
Cross-sectional
surveys
Parents and children
Children are exposed to an overwhelming amount of advertising as
there is little regulation; 30% of Malaysian children watch over 8 h of
television per day during holidays and are exposed to over 2.5 h of
advertisements per day; 73% of Pakistani and 68% of Filipino children
claim to love advertisements.
More than 50% of parents in all countries surveyed said that their
children are an important influencing factor in their purchase decisions.
Of Pakistani children 73% perceive soft drinks to be healthy for frequent
consumption. In the Philippines, 80% of children and 71% of parents
drink soft drinks at least once per week, as do 71% of those in the
Republic of Korea. In the Philippines 40% of parents and 63% of
A-79
No
children believe fast food to be fit for frequent consumption.
Folta, Bourbeau
&
Goldberg,
2008
USA
Qualitative study:
focus groups
A random sample of
(N=52) children from a
typical
urban,
multicultural community
in Massachusetts. 2-4
children per group and
equal numbers of boys
and girls. The average
age was 9.3 years.
Overall, many children understood the purpose of commercials and
were sceptical about claims that were made. All children could
distinguish what is real in a commercial, either in their first reaction or
upon further reflection, and were aware of the persuasive intent.
No
The children judged advertised foods based on a number of factors, not
all nutrition-related. They identified positive and negative nutritional
attributes of foods, but many were equally influenced by messages
about taste and by incentives to buy the product.
Children evaluated both the yogurt snack and high-sugar cereal as
“good and bad” after viewing the advertisements. Sugar or artificial
sugar was the “bad part.” Milk and the free DVD were the “good part” of
yogurt. They were particularly interested in the free DVDs that came
with products.
French
2001
et
al.,
USA
Randomized
controlled
experiment
Users of 55 vending
machines at 24 sites
(12 secondary schools
and 12 workplaces)
Both the pricing strategies and the promotion strategies influenced sales
of low fat snacks. Price reductions were significantly associated with
percentage of low fat snack sales (F(3,66) = 156.89, P < 0.001). With no
price reduction, 10.9% of total sales were for low-fat snacks. This
increased by 9%, 39% and 93% with price reductions of 10%, 25% and
50% respectively (P < 0.05). The number of low fat snacks sold (as
opposed to the percentage) did not differ significantly between the
control and the 10% price reduction condition. There were significant
increases in the absolute number of low fat snack sales in the 25% and
50% price reduction conditions, compared with the other two conditions
(P < 0.05). In other words, the 10% price reduction increased the
percentage of snack sales which were for low fat products without
increasing the absolute number of low fat snacks sold or the total sales
volume, suggesting that customers may have been substituting a low fat
snack for a regular snack. However, with a 25% and 50% reduction, the
absolute number of low fat snacks sold increased, as did the total sales
volume (in the 50% reduction condition). This suggests that customers
increased the number of snacks they bought from the machine, and may
have actually increased their overall calorific intake.
Promotion (labelling and signage) was significantly and independently
associated with increased low fat snack sales (F(2,44) = 3.48, P < 0.04).
The percentages of low-fat snacks sold in the no signage, labelling, and
labelling plus signage conditions were 14.3%, 14.5% and 15.4%
respectively. Only the labelling plus signage condition differed
A-80
Yes
significantly from the no signage condition in post hoc means
comparisons (P < 0.05). The total number of low fat snacks sold did not
differ significantly by promotion condition, suggesting that the
promotions did not increase the total number of low fat snack sales, only
the percentage of snacks sold which were low fat.
There were no significant main effects for price or promotion, and no
significant 2-way interactions, on machine profits, indicating that profits
were not significantly affected by any of the experimental conditions.
Overall sales volume was unrelated to promotion, but was related to
price reduction, with a significant increase in sales volume in the 50%
price reduction condition compared with the other three price conditions,
which did not differ significantly from each other. There were no
differences between adolescents and adults in price sensitivity.
Sales of low fat snacks increased significantly and proportionately with
increasing price reduction, and promotional labels and signage also had
a small, independent effect on low fat snack sales. These effects
occurred in both adult (workplace) and adolescent (school) populations.
Galst & White,
1976
USA
Non-randomized
experiment
plus
observational study
41 children of both
sexes aged 3–11 years
(mean age range 4–7)
and their mothers
Children made an average of 15 PIAs during the supermarket trip,
equivalent to one for every two minutes in the store. Nearly two-thirds
(64%) of the PIAs were independent (i.e. not made in response to a
question from the mother) verbal requests made in front of the item
display on the shelf. The most heavily requested items were cereals and
sugars and confectionery (8% of all requests each), followed by
vegetables (6%), fruit (6%) and articles and plastic goods (6%). Fortyfive per cent of PIAs were successful.
Spearman rank correlation coefficients found a significant positive
relationship between overall television reinforcement value and number
of PIAs made (r = 0.64, P < 0.01), and between the commercial
reinforcement ratio and number of PIAs made (r = 0.52, P < 0.01). In
other words, the more effort a child exerted to keep the overall
videotape playing, and the more effort they exerted to watch the
commercials compared to the programme, the more PIAs they made per
minute in the supermarket. Age was correlated positively with the
commercial reinforcement ratio (r = 0.28, P < 0.05) and the overall
television reinforcement value (r = 0.45, P < 0.01), although this may
have been a study artefact, in that older children may have been more
adept at pressing the response button than younger children. Age was
also correlated positively with total number of PIAs (r = 0.44, P < 0.01)
and number of independent PIAs (r = 0.37, P < 0.01), with older children
A-81
Yes
making more attempts than younger children. The number of PIAs made
correlated positively and significantly with total number of hours of
commercial television watched per week (r = 0.31, P < 0.05), but not
with total number of hours of non-commercial television watched. There
was no significant correlation with age, suggesting that the relationship
between commercial television exposure and PIAs was not a function of
age.
Overall, the study suggested that the more effort a child put into
watching television commercials, as compared with programmes, the
greater the number of attempts to influence mothers’ shopping
purchases he or she made at the supermarket. The fact that only hours
of commercial television watched per week (as opposed to hours of noncommercial television watched per week) also correlated significantly
with number of purchase attempts lent further support to the relationship
between commercials and purchase influence behaviour.
Galst, 1980
USA
Randomized
controlled
experiment
65 children aged 3.5–
6.75 years
One-way analysis of variance on baseline snack choice proportions
indicated that the groups differed in their choices prior to the
intervention. Therefore analysis of co-variance for each of the four
intervention weeks was performed using baseline proportions as the covariate variable. The average proportion of snacks with added sugar
content selected during weeks 3–6 was calculated for each condition,
and Scheffé multiple-contrast tests were performed to locate the
sources of the differences demonstrated by the significant main effect.
Children in the control condition (no adverts) requested significantly
more sugared snacks than children in three of the experimental
conditions: adverts for added sugar foods viewed without adult
comments; adverts for added sugar foods viewed with adult comments;
and adverts for no added sugar foods plus dietary PSAs, viewed with
adult comments. Children who were exposed to the adverts for no
added sugar foods plus dietary PSAs viewed with adult comments,
requested significantly fewer sugared snacks than children in three of
the conditions: adverts for added sugar foods viewed without adult
comments; adverts for added sugar foods viewed with adult comments;
and adverts for no added sugar foods plus dietary PSAs, viewed without
adult comments.
Children exposed to the sugar adverts with comments scored higher
(i.e. were more accurate in their perceptions of which snacks were
‘healthy’ and which contained ‘too much sugar’) than children exposed
to the sugar adverts without comments (F(1,23) = 12.16, P < 0.01),
A-82
Yes
children exposed to the non-sugar adverts with comments (F(1,22) =
3.50, P < 0.07), children exposed to the non-sugar adverts without
comments (F(1,23) = 9.20, P = 0.01) and control children (F(1,22) =
19.41, P < 0.001). Pearson correlation coefficients indicated that there
was no relationship between knowledge scores and proportion of sugar
snacks selected. Note that children selected their snacks in groups, so
there may have been pressure to conform to the group norm.
Ghani
2004
&
Zain,
Goldberg, 1990
Malaysia
Cross-sectional
survey
Canada
Naturalistic
experiment
quasi-
252 Malaysian children
aged 10–12 years from
two Grade A Malaysian
primary schools
144 English-speaking
children
and
331
French-speaking
children aged 9–12
years.
Children’s awareness of television advertising and their influence over
parental purchases were significant predictors of their attitudes towards
television advertising.
No
English-speaking children reported a significantly greater number of
children’s cereals in the home than did French-speaking children (F =
5.51, P < 0.02, English mean = 2.42, French mean = 2.03). A significant
effect for income was found, with low income children purchasing a
greater number of children’s cereals than upper-middle income children
(F = 23.92, P < 0.0001, low income mean = 2.56, upper middle income
mean = 1.38). “Both of these main effects must be interpreted in light of
a significant language by income interaction (F = 3.68, P = 0.05)”.
Newman-Keuls analyses indicated that there was a significant difference
between the two low income groups in number of children’s cereals
purchased, with the English-speaking group purchasing significantly
more (mean = 3.59 vs 1.44, Newman-Keuls P < 0.05), but that there
was no difference between the two upper-middle income groups (mean
= 1.44 vs 1.30, ns).
Yes
Correlation between level of American television viewing and children’s
cereals purchased was 0.35 for the English-speaking sample (P <
0.0001) and 0.19 for the French-speaking sample (P < 0.01). This
suggested that within each language group, purchase of cereals
increased with greater exposure to American television.
American television viewing scores for both English- and
speaking children were divided into three levels, with a third
group in each level (low, medium and high). This independent
was used in one-way ANOVAs, with number of children’s
purchases as the dependent measure.
Frenchof each
variable
cereals
One-way ANOVA for English-speaking children revealed significant
effects (children’s cereals purchased F = 6.90, P < 0.01). NewmanKeuls test revealed that children with the highest level of American
television viewing had significantly more children’s cereals in their
A-83
homes (mean = 3.81) than did children with medium and low levels of
American television viewing (2.23 and 1.23 respectively). Similarly, oneway ANOVA for French-speaking children revealed significant effects for
level of American television viewing on children’s cereals purchased (F
= 5.24, P < 0.01). Newman-Keuls test indicated that cereal purchase
scores for children with highest levels of American television viewing
were significantly higher (mean = 2.66) than cereal purchase scores for
children with medium and low levels of American television viewing
(mean = 1.89 and 1.49 respectively). The relationships within each
group supported the explanation that the differential level of American
television exposure, rather than cultural or income characteristics, was
responsible for at least part of the difference between the two language
groups in purchase of children’s cereals.
It was hypothesized that if English-speaking and French-speaking
children who watched the same amount of American television had
similar scores for cereal purchase, other factors could be ruled out as
independent influences, whereas if they had differing cereal purchase
scores, other independent influences may have been operating on
cereal purchase. The entire sample of children was therefore divided
into low, medium and high levels of viewing of American television. At
each level, a comparison for English- and French-speaking children was
made for the cereal purchase variable. t-tests indicated that none of the
comparisons were significant; this supported the hypothesis that Frenchand English-speaking children viewing the same amount of American
children’s television respond similarly to products advertised during
those programmes. The investigators argued that the lack of differences
lessened the likelihood that other language or cultural factors might
explain the differences in cereal purchase.
To minimize the problem that television viewing was measured more
sensitively (as a continuous variable) whereas language was a
dichotomous variable, level of American television viewing was
dichotomized at the midpoint. A dummy variable regression was
performed, with American television viewing, language and income as
independent variables and cereals purchased as the dependent
variable. This indicated a significant main effect for level of American
television viewed (F = 14.30, P < 0.001), with children in the high level
having purchased more children’s cereals (mean = 2.67) than children in
the low level (mean = 1.62). There was also a significant main effect for
income (F = 19.78, P < 0.0001), with low income children having
purchased more children’s cereals (mean = 2.42) than upper-middle
income children (mean = 2.03). There was no significant effect for
A-84
language, and no significant interactions, although the interaction of
level of American television viewed by income approached significance
(P < 0.07). The difference in children’s cereals purchased as a function
of level of American television viewing was larger for the low income
groups than for the high income groups.
Goldberg, Gorn
&
Gibson,
1978a;
Goldberg, Gorn
& Gibson 1978b;
Study 1
USA
Randomized
controlled
experiment
80 1st grade (5–6-yearold) children in three
‘upper middle class’
schools in Northern
California Bay Area
Children exposed to sugared food adverts selected a significantly
greater number of sugared foods than children exposed to PSAs (12.58
vs 8.70, P value not given). Similarly, children exposed to sugared food
adverts selected a greater number of sugared foods than children in the
control group (12.58 vs 10.20, P < 0.05). Children exposed to the PSAs
selected fewer sugared foods than children in the control group,
although this difference was not significant. There was a significant main
effect on number of sugared foods selected for message type (i.e.
sugared food adverts vs PSAs) (F = 7.47, d.f. = 1.57, P < 0.01). The
same pattern of results was obtained when snack foods and breakfast
foods were analysed separately. There were no significant effects on
number of sugared foods selected for level of exposure (4.5 minutes vs
9 minutes) and no significant interaction.
Yes
To test the hypothesis that repeated exposure might increase the
likelihood of preferences generalizing from the advertised foods to
unadvertised foods in the same category (e.g. from one brand of sweets
to another), the mean number of non-advertised sugared foods was
compared with the number of advertised sugared foods in three different
levels of exposure (9 minutes, 4.5 minutes, control). Children exposed to
the 9 minutes of adverts selected more non-advertised sugared foods
than children exposed to the 4.5 minutes of adverts (P < 0.10) and than
children in the control group (P < 0.05). There were no differences
between any of the groups in the numbers of foods they identified as
healthy and unhealthy.
Goldberg, Gorn
&
Gibson,
1978a;
Goldberg, Gorn
&
Gibson,
1978b;
Study 2
USA
Randomized
controlled
experiment
st
122 1 grade (5–6year-old) children
Children exposed to the Fat Albert programme in all three experimental
conditions selected significantly fewer sugared snacks than children in
the control group (P < 0.05). Children in Condition three (programme
plus sugared food adverts) selected a greater number of sugared foods
than children watching Fat Albert on its own or with nutritional PSAs, but
the difference was not significant.
Children exposed to the Fat Albert programme on its own selected
significantly fewer sugared foods than children exposed in Study 1 to 4.5
minutes of nutritional PSAs (t = 2.91, d.f. = 28, P < 0.01) and to 9
minutes of nutritional PSAs (t = 6.89, d.f. = 28, P < 0.001). There were
A-85
Yes
no differences between any of the groups in the numbers of foods they
identified as healthy and unhealthy.
Gorn
&
Florsheim, 1985
USA
Randomized
controlled
experiment
70 girls
years
aged
9–10
As hypothesized, stronger effects were found for the lipstick condition, in
which exposure to the advertising affected brand and product
preferences, and perceptions of age-appropriate behaviour. Exposure to
the diet drink advertising was associated with increased brand
awareness of diet drinks (P < 0.01), and with increased ability to
complete the jingles in the two diet drink adverts (P < 0.01). Exposure to
the diet drink advertising had no effect on respondents’ personal product
preferences or their preferences if selecting for a teacher, or on their
brand preferences in either of these situations, or on perceptions of ageappropriate behaviour.
Yes
Overall, the study suggested that even where children are exposed to
advertising for a product which may not be salient to them (as diet drinks
were judged to be at the time the study was conducted), the advertising
can increase their brand and advertising awareness and their
perceptions of the link between the product and looking grown-up
Gorn
Goldberg,
1980b; 1982
&
Canada
Randomized
controlled
experiment
288 children aged 5–8
years
There was a significant treatment effect on children’s drinks choices (F
(3,280) = 4.18, P < 0.01). Children exposed to the fruit adverts selected
the most orange juice and children exposed to the sweets adverts
selected the least orange juice (45% vs 25%, P = 0.05). Children
exposed to PSAs and to no messages or adverts fell between the two
other groups (40% and 35% orange juice respectively) and were not
significantly different from the children in the sweets adverts condition.
There was also a significant treatment effect on children’s food choices
(F (3,280) = 5.32, P < 0.001). Children exposed to the sweets adverts
picked significantly less fruit (25%) than children in the other three
groups, which were not significantly different from one another (fruit
adverts 36%, PSAs 35%, control 33%).
There were no significant differences between groups of children in their
expectations of the researchers’ and doctor’s preferences, with
expectations of the researcher’s preferences tending to ‘mostly fruit’ and
of the doctor’s preferences tending to ‘all fruit’. There were no significant
differences between groups of older children in their views on what food
and drinks the camp should provide for the new children. The overall
tendency was in the direction of more sweets than candy, but this did
not differ for the different conditions.
A-86
Yes
Gorn
&
Goldberg, 1980a
Canada
Randomized
controlled
experiment
131 boys aged 8–10
years in Quebec
Children in each condition were reasonably accurate in recalling the
number of adverts they had seen. Correct recognition of the name of the
ice cream increased in children who saw three repetitions of the advert
2
compared with one advert only (76% vs 48%, X = 3.40, d.f. + 1, P <
0.10) and in children who saw three different adverts compared with one
2
advert only (95% vs 48%, X = 8.94, d.f. +1, P < 0.10). Increasing the
number of exposures to five did not improve children’s recall further.
Correct recognition of the number of flavours mentioned in the advert
was relatively high. Of children exposed to only one advert, 62% gave
the correct answer, and there were no significant differences between
the treatment groups.
Yes
Analysis of variance indicated that all experimental conditions had an
effect on children’s brand preference for the advertised brand over other
brands (F = 2.59, d.f. = 5, 105, P < 0.05). Newman-Keuls post hoc
analysis indicated that those who viewed three different adverts had
significantly greater preference for the advertised brand than did those
who viewed only one advert (P < 0.01). The preference scores for all
other experimental groups fell within this range and were not
significantly different from one another.
There were no significant differences between any of the groups with
regard to first choice for a food snack (generic preference). However,
children exposed to five different adverts were significantly more likely to
select ice cream as their second choice (45% made this selection),
compared with 10–15% in the other conditions (P < 0.05).
Increased exposure to the adverts did not increase consumption of the
ice cream, and there was a tendency for those seeing increased
numbers of repetitions to eat fewer ounces of ice cream. This was not
the case for those who viewed increased numbers of different
commercials, where there was no discernible relationship between
quantity of ice cream consumed and number of different adverts.
Removing the influence of the child’s weight and treating it as a
covariate did not alter the results.
Gracey
1996
et
al.,
Australia
Cross-sectional
survey
391 year-11 students
(mean age 15.8 years)
of both sexes from two
public schools and one
private school in Perth,
Australia
The results relevant to the review concern the relationship between
television watching and food knowledge, attitudes and behaviour.
There were significantly lower levels of television watching during the
week among private school students compared with public school
students. Television watching during the week correlated negatively with
nutrition knowledge scores (r = −0.1170, P = 0.028). There were no
A-87
Yes
significant correlations between television watching and body mass
index, fat score or food variety score.
Weekend television viewing was significantly correlated with Kinlay’s fat
score, although no details are reported. Linear regression analyses
were conducted with Kinlay’s fat score and the food variety score as
dependent variables. In each model, independent variables comprised
variables that showed significant univariate relationships with these
variables. Weekend television viewing was one of the independent
variables in the Kinlay’s fat score model, but appears not to have been
included in the model of food variety score. This suggests that weekend
television viewing had a significant univariate relationship with Kinlay’s
fat score but not with the food variety score. However, no details are
given of the strength or significance of the univariate relationships. The
regression models controlled for sex, age and school and all
independent variables appear to have been entered in one step. The
linear regression with Kinlay’s fat score as the dependent variable
showed that, controlling for age, fat score was positively associated with
being male (P < 0.001), drinking alcohol (P < 0.05) and weekend
television viewing (P = 0.0513) and was negatively associated with age
(P < 0.05), self-efficacy (P < 0.001) and influence over food bought at
home (P < 0.05). Thus, a higher level of weekend television viewing was
associated with a higher fat score, although this just approached
significance. The regression model explained 22% of variation in the fat
scores.
Guittard, 2008
France
Review
European data and
reviews
the
UK’s
Ofcom (2004, 2006)
and Hastings (2003)
reports and the USA’s
Institute of Medicine
(2005) report.
There have literally been hundreds of articles published in this area over
the last 3-4 decades, most of which come from the USA. Among food
marketing practices, television advertising was and still is the subject of
most research. This led to a number of global reviews, which have been
used as a basis for policy decision-making since the early 2000s.
In marketing and social sciences, it is very unlikely that one side is
wholly right or wholly wrong. The purpose of this review was not to
weigh one paper against another but to select well-supported arguments
as a foundation for action.
The balance of evidence does support the conclusion that television
advertising has a modest direct effect on children’s food
preferences and choices.
There are promotional effects at brand level, distinct from product
category level but the evidence for promotional effects at the level of
A-88
No
overall diet is thin at best.
The balance of evidence does not support that television advertising (or
food promotion more generally) has larger, indirect effects than other
factors affecting children’s lives. More importantly, there is no evidence
weighing these factors against each other so as to determine their
relative influence.
Advertising is undoubtedly only one in a wide range of factors affecting
children’s food choice, health and obesity. From that, it can be easily
understood that advertising – or television viewing more generally – has
its effect indirectly, mediated alongside and through other variables, as
well as directly.
The evidence that reducing exposure to advertising has beneficial
consequences is also mixed. Research on television advertising also
does not offer straightforward guidance regarding the degree of
restriction, partly because there is no easy translation from amount of
advertising viewed to dietary consequences, and partly because little
research has evaluated the relative importance of food advertising by
comparison with other influences on diet.
Research also provides little guidance regarding the influence of forms
of promotion other than television advertising because this has rarely
been examined, notwithstanding the fast changing array of promotional
strategies, particularly for the internet, games, mobile phone and so on.
Halford
2007
et
al.,
England
Within-subjects,
counterbalanced
design
(food
adverts versus toy
adverts).
Two
testing occasions
separated by two
weeks.
(Same adverts as
Halford
et
al.
7
2004 )
93 children (39 males)
aged 5-7 years from a
UK school. Using BMI
scores, 65 were lean
(NW)
and
28
overweight or obese
(OWOB).
Exposure to the food advertisements produced significant and
substantial increases in food consumption in the group as a whole. The
increase in caloric intake was roughly the same magnitude in the two
weight status groups, 17% (112 kcal) in the NW children and 14% (97
kcal) in the OWOB children, and proves significant in both.
Yes
The level of intake of each food was significantly associated with the
type of food (F(4,88) ¼ 126.467; po0.001), and the effect of advert
condition on the types of foods chosen by the children was near
significant (F(4,88) ¼ 2.265; p ¼ 0.069). Beyond their effects on brand
choice, exposure to food advertisements promoted over-consumption in
5-7 year old children.
7
Effects study which passed the inclusion criteria and demonstrated causality in Hastings G et al. (2006). The Extent, Nature and Effects of Food Promotion to Children: A
Review of the Evidence. Technical Paper prepared for the World Health Organization. Geneva, World Health Organization.
A-89
Halford
2008a
et
al.,
England
Within-subjects,
counterbalanced
design (toy advert
control
condition
versus food advert
experimental
condition).
Two
testing
sessions
held at same time
(2
hours
after
lunch) on test days
two weeks apart.
37 children aged 11-13
years from a school in
Liverpool, UK.
24 were normal-weight
(NW),
10
were
overweight (OW) and 3
obese (OB).
et
al.,
England
Within-subjects,
counterbalanced
design (food advert
versus toy advert).
The children were
tested
on
two
occasions
separated by two
weeks.
Yes
No significant differences were found for the OWOB group between the
two conditions, however OWOB children showed a greater preference
for branded foods than NW children per se. In the OWOB group only,
there was a significant positive correlation between food advertisement
recall and the total number of food items chosen in the experimental
(food advertisement) condition (r=0.497, N=13, p=0.042, one-tailed).
OWOB children did not recall more food advertisements than NW
children.
(Same adverts as
Halford
et
al
2008b)
Halford
2008b
Exposure to food advertisements significantly increased the preference
for branded food items (t(23)=2.795, p=0.01) and non-branded food
items (t(23)=2.426, p=0.024) compared with the control condition in NW
children.
59 children (32 male)
aged
9-11
years
recruited from a UK
school.
33 were normal-weight
(NW), 15 overweight
(OW) and 11 obese
(OB).
Exposure to food adverts produced substantial and significant increases
in energy intake in all children (P < 0·001). The increase in intake was
largest in the OB children (P = 0·04). All children increased their
consumption of high-fat and/or sweet energy-dense snacks in response
to the adverts (P < 0·001). In the food advert condition, total intake and
the intake of these specific snack items correlated with the children’s
modified age- and gender-specific body mass index score.
Yes
There was a significant three-way interaction of food consumption
between food type, advert condition and weight status (F(8,108) =
3.429; P=0.001). Using planned post hoc one-way ANOVAs, it as
shown that this significant interaction was due to differences in food
intake between the weight status groups. The OB group ate significantly
more chocolate buttons and Snack-a-Jacks following the food adverts
than both the OW and NW children, and more crisps than the NW
children (all tests p<0.05).
(Same adverts as
Halford
et
al
2008a)
In response to the food adverts, the OB children increased their intake
by 1.97MJ (471 kcal) from the control (toy advert) condition compared
with 1.28MJ (306 kcal) in the OW and 1.05 MJ (250 kcal) in the NW
children.
Halford
2004
et
al.,
United
Kingdom
Repeated
measures design
42 lean, overweight
and
obese
school
children in Liverpool
No significant difference in the number of non-food adverts recognized
by the lean and obese children was observed, but the obese children did
recognize significantly more food adverts. Ability to recognize the food
adverts significantly correlated with the amount of food eaten after
A-90
Yes
exposure to them. Overall snack food intake of the obese and
overweight was significantly higher than the lean children in the control
(non-food advert) condition. The consumption of all the food offered
increased post food advert with the exception of the low fat savoury
snack. These data demonstrate obese children have heightened
alertness to food related cues. Exposure to food adverts promotes
consumption.
Harrison
&
Marske, 2005
USA
Survey at two time
points
132 children in school
grades 1–3
Television viewing predicted subsequent reductions in nutritional
knowledge and reasoning but these findings were significant only for
foods that tend to be heavily promoted as “weight loss aids”. Concludes
that television's framing of diet foods may confuse children by equating
weight-loss benefits with nutritional benefits.
Yes
Heslop & Ryans,
1980
Canada
Randomized
controlled
experiment
280 children aged 4–8
years
in
London,
Ontario,
and
their
mothers
Exposure to the experimental adverts had a significant impact on only
one of the measures; the child’s stated cereal. Preference (P = 0.06)
when the F ratio for the control group versus all other groups was
examined. There was no significant difference between the control and
experimental respondents on the other two measures of preference and
behaviour. The degree of emphasis placed on the free gift in the advert
(i.e. whether all or only part of the advert featured the free gift) had no
significant impact on any of the three measures.
Yes
The proportion of respondents selecting the advertised cereal over
another cereal was always higher for the group exposed to three
repetitions of the advert compared with those exposed to only one, but
the difference was not significant. When compared by age, older
children were significantly more likely to take home the advertised cereal
than younger children.
Hitchings
&
Moynihan, 1998
United
Kingdom
Interviews
children’s
completion
intake diary
plus
selffood
44 children aged 9–10
years (mean age 10.2)
of both sexes
Overall, the study suggested that effects of the advertising and free gifts
on behaviour were minimal, although there was an effect on child’s
preference.
The mean number of adverts recalled in each category ranged from
0.36 for cakes and biscuits to 3.25 for breakfast cereals. Children also
recalled an average of 2.14 confectionery adverts and 1.59 soft drink
adverts. Spearman’s rank correlation analysis indicated that the
strongest relationships between adverts remembered and foods
consumed were for soft drinks (r = 0.68, P < 0.001), crisps and savoury
snacks (r = 0.61, P < 0.001), cakes (r = 0.57, P < 0.001) and sweets (r =
0.56, P < 0.001). All other relationships were also significant apart from
the relationship for chips.
A-91
No
Comparing the 10 most frequently recalled food adverts by children and
the 10 most frequently requested foods as reported by parents, four
items appeared on both lists: Walkers crisps, Kellogg’s Coco Pops,
Micro chips, and Kellogg’s Frosties. Parents reported granting 96% of
children’s food requests.
Jeffrey,
McLellarn & Fox,
1982 Study 1
USA
Exploratory
randomized
controlled
experiment
47 children aged 4–5
years
The analysis of variance with repeated measures (3 x 2 factorial
ANOVA) found significant increases between baseline and post-test in
total calories consumed from foods, (F(2,44) = 9.75, P < 0.01), total
calories consumed from beverages, (F(2,44) = 8.12, P < 0.01) and total
calories consumed from food and beverages (F(2,44) = 14.99, P <
0.01). However, as no significant difference between groups and
interaction effects were found the study did not provide evidence of low
nutrition adverts exerting an influence on the children’s food
consumption.
Yes
Pre–post changes in the three calorific consumption scores were
analysed separately for each of the treatment groups. Comparing
baseline and post-test scores for each group separately, the group
exposed to ‘low nutrition’ adverts increased their total calorific
consumption for foods (t(1,15) = 2.41, P < 0.05), drinks (t(1,15) = 2.67,
P < 0.05) and foods and drinks combined (t(1,15) = 3.38, P < 0.01). The
group exposed to ‘pro-nutrition’ adverts increased their total calorific
consumption only for drinks (t(1,14) = 2.36, P < 0.05). The control group
displayed no significant changes between baseline and post-test.
Calorific consumption scores for the specifically advertised foods were
also analysed separately. Subjects exposed to the ‘low nutrition’ adverts
increased their consumption of two of the three advertised ‘low nutrition’
foods between baseline and post-test (Hersheys Chocolate, t(1,15) =
2.63, P < 0.05; Fritos, t(1,15) = 2.16, P < 0.05). Subjects exposed to the
‘pro-nutrition’ and control adverts did not increase their consumption of
these specific foods.
These separate analyses of pre–post change within each treatment
group were suggestive of a possible influence of ‘low nutrition’ adverts
on the children’s food consumption. However, in the absence of
between groups and interaction effects from the ANOVA, no effect could
be concluded.
Jeffrey,
McLellarn & Fox,
1982 Study 2;
USA
Randomized
controlled pre- and
post-test
Groups of 48 4–5-yearolds and 48 9–10-yearolds, with an equal
There was a significant change x group x sex interaction on the total
calories foods and beverages variable, F(2,84) = 3.60, P = 0.032.
Newman-Keuls test on the adjusted means indicated that boys exposed
A-92
Yes
Fox, 1981
Jones et al.,
2007, Study 2
experiment
Australia
Qualitative study:
focus groups
number of boys and
girls in each age group.
In May 2007, 24 5-9
year olds (12 males)
recruited in friendship
pairs took part 4 singlesex
focus
groups.
Thirty 12-14 year olds
(15 males) took part in
4 single-sex focus
groups. Respondents
recruited from bluecollar (e.g. Parramatta
and Blacktown) and
middle
class
(e.g.
Bauklham Hills and
Castle Hill) areas in
Sydney
by
a
commercial recruitment
company.
to the low nutrition adverts were the only group to display a significant
increase in consumption (P < 0.05). Males in the low nutrition adverts
groups did not differ from males in the other two groups at baseline, but
consumed significantly more food on the post-test than males in the
other two groups (P < 0.05). The change x group x sex interaction on
the low nutrition foods and beverages variable was not significant at the
traditional 0.05 level (F(2,84) = 2.75, P = 0.07). Post hoc comparisons
performed on the adjusted means revealed the same pattern of results
as found on the total calories food and beverage variable, with the low
nutrition advert group for males eating more after exposure to the
television adverts than any other group. In comparison with the other
groups, males exposed to the low nutrition adverts consumed
significantly more calories from all foods at post-test. They also
consumed more calories from low nutrition foods than did other groups,
but this did not reach statistical significance at the traditional P < 0.05
level.
Change x group x sex x age interactions were not significant, suggesting
that the two different age groups were not affected differentially by the
adverts. On the cognitive measures, 9-year-olds scored more highly
than 4-year-olds on definitions of a balanced diet and ability to identify
healthy and unhealthy food, and on recall of the advertised products.
They were also better able to define the difference between
programmes and adverts, and were less likely than 4-year-olds to
believe that programmes and adverts always tell the truth.
Media exposure: Media use for 5-9 year olds involved TV and movies;
only half were permitted to use internet at home (boys more than girls,
mostly to access games), boys media use was predominantly screenbased, and some children identified reading children’s magazines from
the supermarket but could not name titles. A range of media were used
by 12-14 year olds: TV, movies, radio, using the computer for social
networking and downloading music. Boys indicated playing videogames
and only girls indicated reading magazines.
The discussions illustrated how the marketing strategies used by food
companies influence children’s knowledge, attitudes and behaviours in
relation to the advertised foods. Both the 5-9s and the 12-14s showed
that their food beliefs and food preferences were being influenced by
food companies’ marketing strategies in Australia. In relation to food
beliefs, of particular significance is the predominance of non-core food
groups as favourite foods (the ads for Coco Pops/Coco Rocks, and
Cadbury chocolate were the most commonly recalled and liked) and the
association between the playing of sport and the need to consume high
A-93
No
sugar foods for energy. There was reasonable knowledge of healthy
eating practices, particularly in the 12-14s. However, in all groups,
knowledge of healthy eating appeared to have only a small influence on
food preferences. Instead, preferences appeared to be particularly
influenced by the power of the brand, particularly branding with
celebrities (sport and other), cartoon or movie characters. Competitions,
toys and prizes were significant motivators of food purchase.
McDonald’s was overwhelmingly identified as a favourite restaurant,
with toy offers appearing to be integral to this preference. Some
teenagers were critical of marketing strategies utilised by food
companies, however, it is unclear whether this critical perspective
altered food preference or consumption.
Kaufman
&
Sandman, 1983
USA
Randomized
controlled pre- and
post-test
experiment
1108 children aged 5–
10 years from public
schools in three cities
in the USA. Each class
was randomly assigned
to
one
of
four
treatments. A total of
1179
children
completed all parts of
the study, but 71 were
excluded
from
the
analysis because of
“invalid
answer
sheets”,
leaving
a
sample of 1108. The
sample was evenly
distributed by sex and
age
‘Healthy Choice Scores’ at post-test were subject to analysis of
covariance by experimental condition, controlling for both city and pretest scores. This measured the effect of the different experimental
adverts on children’s preferences when the confounding effects of city
and pre-test preferences were statistically eliminated. Both city and pretest scores significantly affected post-test ‘healthy choice scores’ (city F
= 16.6, pre-test ‘healthy choice score’ F = 808.7; P < 0.01). There was
also a significant main effect on post-test ‘healthy choice scores’ (F =
24.0, P < 0.01) when these influences were statistically controlled for.
The explained variance for the analysis was significant at P < 0.01,
suggesting that the relationship between treatment and post-test
‘healthy choice score’ was unlikely to be due to other variables.
Adjusting for pre-test preferences and differences between cities,
children exposed to the sugared food adverts only made fewer ‘healthy
food choices’ at post-test than respondents in other conditions (adjusted
mean 49.76), and children exposed to the counter-adverts (whether
accompanied by sugared food adverts or not) made the most ‘healthy
food choices’ (counter-adverts alone 61.40, counter-adverts plus
sugared food adverts 60.93); children exposed to the sugared food
adverts with disclaimers fell between these groups (56.69). No control
was made for possible age variation.
The investigators suggest that post-test scores may have been
influenced in either of two directions by the research design. Children
may have perceived the post-test task as altering their preferences in
the direction of the adverts they had just seen, and the post-test scores
may therefore have represented a willingness to alter their responses to
match the adverts: i.e. the differences between the groups may have
been enlarged by a willingness to comply with the perceived purpose of
A-94
Yes
the test. Alternatively, the task of completing the pre-test may have
“initiated a strain toward consistency for some children”, deterring them
from changing their responses after seeing the adverts in order to
maintain consistency. This may have diminished differences between
groups.
Lam, 1978
USA
Cross-sectional
survey
44
children
aged
between 4 and 7 years,
and their mothers
Twenty-five per cent of children said that they ‘always’, and 59% that
they sometimes, demanded cereal products that they saw advertised on
television. Mothers’ reports of children’s demands fell within a similar
range (18% always and 75% sometimes). Mothers described yielding to
children’s demands most of the time (9%), sometimes (55%), very
seldom (25%) and never (11%).
No
Boys reported that their favourite adverts were for toys (62%), food
(13%) and cars (13%), while for girls the favourites were toys (39%),
cereals (14%), food (12%) and dolls (12%). Food and cereal adverts did
not feature in the list of least favourite adverts.
Lewis
1998
&
Hill,
United
Kingdom
Non-randomized
experiment
103 children (51 girls,
52 boys) aged 9–10 in
Year 5 from two state
schools in the north of
England. The sample
represented 90% of the
school registers for the
year
group;
the
remainder were absent
on one or more of the
research days.
The
sample is described as
majority white (82%)
and from a low to
middle class social
background.
No
information is provided
on how the two schools
were selected.
The group of overweight children had a significantly greater preference
to be thinner than the ‘normal weight’ children (P < 0.001) and were
significantly less satisfied with their physical appearance (P < 0.01).
Of the ‘current state’ ratings, only one, ‘feeling fat’, was significantly
related to children’s weight, with overweight children feeling significantly
more fat than normal weight children (P < 0.001). Viewing the adverts
had significant effects on mood, with children reporting feeling
significantly less worried and less sad, and more liked by their friends,
after exposure to the adverts (P < 0.05 for all three measures). There
were no significant main effects by advert type, although there were
significant interactions on two rating scales ‘feeling healthy’ and ‘feeling
like eating sweets’.
There were three-way interactions between advert type, time and
weight. Thus, overweight and normal weight children responded
differently to the two advert types on their ratings of ‘feeling healthy’
(three-way interaction: F (1,33) = 11.26, P < 0.01) and desire to eat
sweets (three-way interaction: F(1,33) = 5.80, P < 0.05). After viewing
the food adverts, overweight children felt healthier and felt less like
eating sweets while comparison children felt less healthy and more like
eating sweets. The opposite occurred after viewing the non-food
adverts, with overweight children reporting that they felt less healthy and
more like eating sweets, while normal weight children reported feeling
more healthy and had less desire to eat sweets.
A-95
Yes
Lobstein & Dibb,
2005
USA, Australia
& 8 European
countries
Surveys
countries
Marshall,
O'Donohoe
Kline, 2007
New Zealand data
Secondary analysis
of
survey
and
qualitative data
&
Scotland
Canada
analysis
in
13
Various
A significant association was found between the proportion of children
overweight and the numbers of advertisements per hour on children’s
television, especially those adverts that encourage the consumption of
energy-dense, micro-nutrient-poor foods, (r = 0.81, P < 0.005). A weaker
negative association was found between the proportion of children
overweight and the number of adverts encouraging healthy diets (r =
−0.56, P < 0.10). The quantity of advertising on children’s television
appears to be related to the prevalence of excess body weight among
children.
No
8-11 year old children
from New Zealand
HFSS food ads were well-represented in their repertoire of favourite
ads, and they reported being influenced by these. However, their
accounts of snacking highlighted the extent to which their actual
consumption was shaped by parental agendas and concerns.
No
&
-
Although they gravitated towards less healthy snack foods, fruit, and
vegetables were included in their categorization and repertoire of
snacks, perhaps reflecting the level of monitoring and gatekeeping
exerted by their parents, who established ground rules for snacking and
in many cases directly controlled their access to snack foods, although
the limits imposed varied according to context. The children were
generally accepting of this, although they drew on a range of strategies
and tactics to access their preferred snacks.
In the survey 67% reported that food ads ‘often’ or ‘sometimes’ made
them feel hungry, 78% that seeing an advertisement on television made
them want to buy the product ‘sometimes’ or ‘often’.
Maryam et al.,
2005
Islamic
Republic
Iran
Maskill
1996
New Zealand
et
al.,
of
Cross-sectional
survey
Qualitative
interviews
adolescents
carers;
survey
adolescents
with
and
of
398 junior high school
students from six state
and six private schools
in a middle class
district of Tehran
20 Pakeha (European)
families with 13−16year-olds
“Ashimashi’s” puffed cereals was the most recalled advert. Food adverts
are not based on sound nutritional principles and are therefore
misleading. These cereals were shown in the content analysis part of
the study to be the largest category of advertised foods (36%).
No
Not many of the young people perceived television advertising to have a
significant effect on their purchasing or consumption behaviour,
however, it was perceived to have an effect on some teenagers’
decisions to buy certain product categories (sweets, takeaways and
crisps). Television advertising appeared to influence young people by
raising their awareness of a particular food (often new foods or those on
special offer). Several respondents also reported that television
advertising had some influence on them wanting to eat more takeaways,
sweets, breakfast cereals, bread, fruit and crisps.
No
A-96
Common reasons for liking food advertising included that they were
‘cool, humorous, far-fetched, had catchy tunes, made you aware of the
food and made you feel hungry.’
Matheson et al.,
2004
USA
Three consecutive
24h dietary recalls
collected from each
child
Ethnically diverse thirdgrade children and
predominantly Latino
fifth-grade children
On weekdays and weekends, 17–18% and approximately 26% of total
daily energy were consumed during television viewing in the two
samples, although the fat content of foods consumed during television
viewing did not differ significantly from that of the foods consumed with
the television off; less soda, fast-food, and fruit and vegetables were
consumed with the television on. The amount of food consumed during
rd
television viewing was not associated with children’s BMI, but in the 3
grade sample the fat content of foods consumed during television
viewing was associated with BMI.
Yes
Moya
De
Sifontes
&
Dehollain, 1986
Venezuela
Cross-sectional
survey
263 housewives (or the
person who fulfilled the
role of housewife); 8%
were upper class, 31%
were middle class, and
61% were lower class
A preliminary content analysis of all adverts for food products featured in
the mass media (television, radio and the press) during one week. The
exercise provided researchers with some background data on the types
of food advertising going on, and informed the development of the
questionnaire.
No
Use of social mass media: Nearly all homes (98%) had a television (the
2% that did not were all lower income homes). Three-quarters of
mothers said that they watched television with their children, usually for
2–4 hours per day. Nearly all children watch television (99%). Children’s
viewing habits (in terms of frequency of viewing and preferences for
different television channels) were consistent with the viewing habits
and preferences of their mothers, and were also consistent across
different social class groups.
Nearly half of mothers (44%) listened to the radio for < 2 hours per day.
Over a quarter (27%) listened for an average of 2–4 hours, 16% listened
for more than 4 hours, and 13% did not listen to the radio. There was a
slightly positive relationship (gamma coefficient of 0.20, P < 0.05)
between the amount of time spent listening to the radio and social class:
upper class mothers tended to listen to the radio for longer periods that
lower class mothers (2–4 hours on average per day vs 2 hours
respectively).
Nearly all mothers (97%) read the press; 68% read print media on a
daily basis (this was consistent across social class groups). The authors
also note that the preliminary content analysis of newspapers indicated
that three newspapers (‘El Nacional’, ‘El Universal’ and ‘Ultimas
Noticias’) contained more adverts for commercial food products, and
A-97
that these newspapers were indeed read by mothers included in the
study.
Recall of food advertising: Nearly all mothers (97%) said that they could
remember some form of food advertising featured in the mass media.
No statistically significant differences were observed across social class
groups in this respect, although, in purely descriptive terms, mothers in
middle class and lower class groups were better able to remember food
advertising than upper class mothers (99% vs 97% vs 90%
respectively).
There were differences in the types of food adverts recalled by mothers
in different social class groups. High protein food adverts were well
remembered among upper class mothers and cereal advertising was
better remembered among mothers from low and middle social class
groups.
Liking for and attitudes towards food advertising: Over half of the sample
(52%) indicated a preference among children for food adverts.
Differences by social class were non-significant (X2 = 3.14, P = 0.05).
Over 70% (73%) of mothers considered television as the primary media
influence on their children’s preferences for advertised foods, just 1%
felt that either the radio or the press was the most significant influence,
and a quarter of respondents said that they ‘did not know.’
There was a significant relationship (P = 0.01) between social class, and
mothers' beliefs about the importance of highly advertised commercial
foods in their children’s diets. Lower class mothers attributed greater
importance to the commercial foods advertised in their children’s diets
(gamma coefficient = 0.30). Cereals and jams were considered by
mothers as among the top three most important foods in children’s diets,
followed by jelly and ice cream. Little importance was attributed to fizzy
drinks, despite the fact that fizzy juice is frequently consumed by
children.
Mothers were questioned directly about the extent to which they
believed food advertising messages. Again, a significant relationship
was observed between mothers’ beliefs and social class. As social class
decreases, the proportion of mothers influenced by advertising
increases (gamma coefficient = 0.32).
Preferences for different foods: Most children (consistent across social
A-98
class groups) demonstrated preferences for chocolate drinks, cereals,
jelly, pork, sausages and ice cream (in that order of importance). There
was a significant relationship (P = 0.05) and a positive correlation
(gamma coefficient 0.28) between time children spend watching
television and their ‘preference for commercial foods broadcast by the
social mass media’.
Purchase-related behaviour: A significant relationship was also
observed between mothers’ responsiveness to children’s requests for
advertised food products and social class; lower social class mothers
demonstrated a greater tendency to respond to children’s requests (P <
0.05).
Food purchase behaviour: Nearly half of all mothers (44%) purchased
foods from supermarkets. The relationship between social class and
choice of purchasing outlet was statistically significant (P < 0.05).
Seventy-seven per cent of upper class mothers and 79% of middle class
mothers bought foods from supermarkets (compared with only twothirds of lower class mothers).
Just over half (52%) of mothers shopped for food alone, and 30% went
shopping with their husband. Children were not significantly involved in
food shopping: 9% of mothers went shopping with older children, and
6% typically went shopping with their younger children. No significant
social class differences were observed.
By way of establishing whether mothers consciously bought advertised
foods, they were asked to provide reasons for their food purchase
choices. Less then half of mothers confirmed that they had purchased
products they had seen or heard advertised on television, radio or in the
press. No social class differences were observed.
Mothers were also questioned more specifically about purchases made
for food products that had been advertised during the course of the
research. In this respect, lower class mothers demonstrated a greater
tendency to purchase advertised foods: 46% vs 37% of middle class
mothers and 33% of upper class mothers. Upper class mothers
demonstrated a tendency to purchase high protein foods, compared with
lower class mothers who demonstrated a tendency to purchase
carbohydrates. Cereals and chocolate drinks were the most frequently
purchased products. This effect was consistent across all social class
groups.
A-99
Food consumption behaviour: Jams and fizzy drinks were frequently
consumed among lower social class children, and jams and cereals
were most consumed by middle class children. Preferences among
upper class children were for fizzy drinks and yoghurts. The authors
note that all these products are sweetened foods.
Overall differences by mothers' occupation and level of education were
not significant. However significant differences in age were observed;
older mothers (> 39 years) showed less of a tendency to believe in the
‘food advertising process’.
Musaiger et al.,
1986
Bahrain
Cross-sectional
survey
Mothers from 1260
households in Manama
City (capital of Bahrain)
Hours children spend watching television: Over 60% of all children
watch more than 3 hours of television per day. The association between
television viewing and social class was statistically significant, as
socioeconomic status increased, children spent less time watching
2
television (X = 134.59, P < 0.001).
Children’s favourite food adverts: Over half of all children (52.8%)
favoured advertising for ‘candies and chocolate’. This finding was
consistent across low, (52.7%) middle (51.3%) and high (56.7%) social
class groups. This was followed by advertising for ‘milk and milk
products’ (28.7%) of all children, baby foods (7.1%), frozen foods
(1.6%), coffee and tea (0.8%), and others (9.0%).
Children’s requests for advertised foods: Nearly 60% (59%) of all
children ‘always’ request foods advertised on television, 29.8%
‘sometimes’ and 11.2% ‘rarely’. The relationship between social class
and requests for advertised foods was statistically significant (X 2/4 =
50.29, P < 0.001), as lower and middle social class children were more
likely to request advertised foods than upper social class children. Twothirds (66.2%) of lower social class children ‘always’ request foods
advertised on television, 52.7% of middle class children ‘always’
request, compared with 46.6% of upper class children. [X 2/4 = 50.29 (P
< 0.001)].
Children aged between 6 and 12 years demonstrated a greater
tendency to request advertised foods (53.2% of entire sample) than
children under 6 (31.1%) or over 12 years (15.7%). This trend was
particularly evident among lower social class children (58.1%, compared
with 47.9% and 43.3% of middle and upper class children respectively)
[X2 = 91.35, (P < 0.001)].
Mothers’ responses to children’s requests for advertised foods: Over
A-100
No
60% of all mothers (61.1%) were ‘always’ responsive to children’s
requests for advertised food, 34.4% were ‘sometimes’ responsive and
only 4.1% ‘do not respond’. Again, the relationship between mothers’
responsiveness to requests and social class was statistically significant
2
[X = 41.91 (P < 0.001)]. Again, lower income groups were more
responsive: 67.3% of lower class mothers were ‘always’ responsive to
children’s requests, compared with 58.2% of middle class mothers and
41.6% of upper class mothers.
Neeley
&
Schumann, 2004
USA
Post-test between
subjects
factorial
designs
2–5-year-olds
Although character action and voice may influence attention to the
advert, character and product recognition, and a positive attitude
towards the product, the relationship between preference, intention and
product choice remained unclear.
Yes
Norton, Falciglia
& Ricketts, 2000
USA
Survey
35 adolescents aged
9–18 years (19 males,
16
females).
All
subjects were white
and middle class. All
respondents were one
member of a pair of
twins
Overall, the foods perceived by the researchers as more frequently
eaten by young people were also the most preferred foods, with
spaghetti, cola, sugared cereal, apples, snack cake, French fries,
orange juice, chicken and hamburger all receiving ratings of 6.74 and
above.
Yes
Overall, taste was significantly correlated with preference for the highest
number of foods (15 of the 17 foods). ‘Healthfulness’ was significantly
related with preference for four of the foods, as was accessibility
(although not with the same four foods). ‘Peers eat it’ was significantly
correlated with preference for three of the foods. ‘Parents serve it’ was
significantly correlated with preference for one of the foods (broccoli), as
was advertising (chicken). Price was not significantly correlated with
preference of any food.
Stepwise regression of the seven motivational factors indicated that
taste had the greatest impact on food preference, followed by
advertising, peers eating, parents serving, accessibility and
healthfulness. Price did not influence preference for any item. Taste was
a significant influence on 16 food items, while advertising was a
significant influence on three (apples, beans, low fat milk).
‘Healthfulness’ was an influence only on unsweetened cereals. Parents
serving was an influence on cheese and whole fat milk, and peers
eating was an influence on apples and chicken. Accessibility was an
influence on unsweetened cereal and whole fat milk.
No standardized regression coefficients were reported so it is not
possible to assess, from this multivariate analysis, the strength of the
independent influence of television advertising relative to other
A-101
significant variables. However, the analysis does show that television
advertising (P < 0.05), taste (P < 0.01) and ‘peers eat it’ (P < 0.05) all
significantly and independently influenced preferences for apples while
the remaining motivational factors were controlled for. Similarly,
television advertising (P < 0.05) and taste (P < 0.01) significantly and
independently influenced preference for beans, and television
advertising (P < 0.01) and taste (P < 0.01) significantly and
independently influenced preference for low fat milk. These latter two
regressions also controlled for the remaining motivational variables that
were not found to be significant. While the regression coefficients are
not available, to judge the relative influence of television adverts
compared with other motivational factors, the strength of the Pearson
correlation coefficients and the finding that taste influenced so many
preferences together suggest that taste had a stronger influence than
television advertising.
Olivares et al.,
1999
Chile
Cross-sectional
survey
786
school-aged
children aged 6–11
years (44% boys, 56%
girls).
Three socioeconomic levels were
represented:
Low
and low
or
average level (43% of
schools)
Mean level (38.3%)
High and high mean
level (18.7%)
Television viewing behaviour (as a proxy measure of food advertising
exposure):
Nearly all children (99%) said that they watched television at home
(independent of age, sex, or socioeconomic status). Over 20% of
children watch more than 3 hours of television on weekdays and over
40% watch 1–3 hours of television. Significantly more 6–8-year-olds
than 9–11-year-olds watch more than 3 hours of television per day
(23.1% vs 13.9%). No significant differences according to sex or
socioeconomic status were observed.
Liking for food advertising: 80% of all children could recall food or drinks
adverts that they liked. This effect was greater among older children (9–
11 years). The three most popular food and drink adverts were for
snacks (32.6%), drinks (32.5%), and yoghurt and milk (11.7%).
Preferences for milk/yoghurt adverts were greater among children from
low or average social class groups. Advertising for other product
categories mentioned includes: sauces, mayonnaises, ready meals and
other non-nourishing’ products.
Preferences for different foods: ‘In accordance with their favourite
adverts’, the children were also questioned about their favourite foods.
The most popular food type was snacks, listed by over half of all
children (56.0%), and followed by ready meals (22.1%) and home-made
meals (21.9%). Fizzy juice was the favourite type of drink (74.2%).
Influence of advertising on food-related behaviour: 70% of students
expressed an interest in trying new food or drinks advertised on
A-102
No
television. When asked if they had consumed food or drinks that had
appeared in adverts of the previous day, half (50.6%) of 6–8-year-olds
and 66.5% of 9–11-year-olds said that they had consumed at least one
product.
Over 80% (83.8%) of all children said that they had money to buy food
and drinks (there were no significant differences across socioeconomic
groups). Children were asked about the foods that they bought: twothirds (66.6%) of all children bought snack products, 14.5% bought fizzy
drinks or juice, and 6.9% bought yoghurt or milk. The authors comment
that the children’s purchase choices are consistent with their advertising
preferences (snack products always the first choice). Gender analysis
revealed significant differences between boys and girls: more girls than
boys bought snacks, and more boys than girls bought drinks. Threequarters (74.8%) of all children said that they bought food or drinks
advertised on television with offers of prizes or gifts: this effect was
significantly greater among children in low (75.2%) and average (78.3%)
socioeconomic groups (vs 66.4% of children from high socioeconomic
groups). This effect was also greater among children aged 6–8 years
(77.3%) than 9–11 years (70.0%). Furthermore, 64.9% of children said
that they continue to buy such products even when the offer of a prize or
gift has ended.
Consumption (eating and snacking) behaviour: Children were also
asked about their general eating habits. Eighty per cent of all children
took a snack to school. A greater proportion of students of high
socioeconomic status brought snacks to school. Lower social class
children also reportedly took more fruit to school (36.6%) than average
(27.9%) and high (14.9%) social class children. Overall, 96.9% of
children ate homemade food in school and at lunch time, only 3.1% ate
fast-food or another commercial product (although 8.4% of high social
class children ate fast-foods). Over 70% (71.6 %) of all children said
they ate foods outside ‘habitual eating hours’. Snacks were the most
popular choice of foods among children at these times (54.3%), followed
by drinks and juice (20.1%); 14.7% ate yoghurt or milk products and
10.9% plain bread or with an accompaniment. No significant differences
in terms of sex, age or socioeconomic status were observed.
Olivares, Yáñez
& Díaz, 2003
Chile
Cross-sectional
survey
274
primary-school
children (aged 10–13
years) from Copiapo,
north Chile (31.8% of
children),
Melipilla,
Television viewing behaviour (as a proxy measure of food advertising
exposure): 91.9% of all children watched television every day
(differences by city or gender were not significant). A quarter of all
children (24.7%) watch more than 5 hours of television on weekdays
and nearly one third (31.8%) watch 3–4 hours. Children reportedly
A-103
No
central Chile (37.9%),
and Delcahue, south
Chile (30.3%)
watch more television on Saturdays: 27.5% watch more than 5 hours
and 30.9% watch 3–4 hours (again, differences by city were not
significant).
Preferences for food advertising: Nearly 70% of children said that they
liked to watch television adverts and nearly 90% (88.7%) remembered
food adverts that they liked (differences by city or gender were not
significant).
Almost 40% (38.5%) of all children said that they preferred adverts for
sweet and salty products (including French fries, salty snacks,
chocolate, cakes and other products saturated in fat, sugar and salt, and
low in fibre. (As much as 46.8% of children in Copiapo preferred
advertising for sweet and salty products). Of all children, 28.4%
preferred adverts for drinks/refreshments, 19.5% preferred yoghurt
adverts and 13.6% preferred fast-food advertising.
Influence of television food advertising on food consumption and
purchase behaviour: 40% of all children said that they had consumed
some food or drink that had featured in advertising of the previous day.
Over 70% of children showed an interest to try new foods promoted
through television advertising.
Children were also asked about their own food purchasing behaviour.
Over a third of all children (34.3%) said that they ‘always’ had money to
buy the foods and drinks they wished, and over 60% (64.2%) said that
they ‘sometimes’ had the money to buy what they wanted. (Nearly half
of children in Melipilla ‘always’ had money to spend as they wished).
Children’s first priority of purchase was sweet and salty food products
(68.1%); 17.2% bought drinks; only 6.3% bought yoghurt/ milk; and only
4.7% bought other products. Children in Delcahue demonstrated a
greater tendency to buy sweet/salty products than children in the other
two cities.
Children were also asked about the foods they brought to school. Nearly
40% (38.7%) of all children brought a lunch from home, of which nearly
half (47.4%) consisted of sweet and salty products (this figure was as
high as 60.4% in Delcahue). Only 19.5% of children brought fruit and
18.1% yoghurt. Differences by city were not significant.
Pavlu, 2001
Czech
Republic
Cross-sectional
survey
121
mothers
of
schoolchildren aged 8–
11 years
Over 80% (81%) of mothers felt that advertising had some influence on
their child; 82% said that their children, under the influence of television
adverts, request them to buy certain products. Most often, these
A-104
No
Peterson et al.,
1984
USA
Randomized
controlled pre- and
post-test
experiment
106
‘kindergarten’
grade children of both
sexes aged 5–6 years,
with a mean age of 6.2
years.
requests are for drinks, candies, toys and dairy products.
On the first set of nutritional knowledge questions, both the experimental
and control groups scored higher at post-test. There was a significant
main effect for trials, F(1,4) = 10.13, P < 0.05, but no significant
treatment by trials interaction was obtained. On the second set of
nutritional knowledge questions, designed to test learning of the specific
nutritional concepts presented in the tapes, a significant main effect for
trials was obtained, F(1,4) = 55.00, P < 0.001. A significant treatment by
trials interaction was also obtained, F(1,4) = 24.48, P < 0.01.
Yes
Repeated measures ANOVA performed on the food preferences scores
revealed significant main effects for trials on both composite preferences
for ‘pro-nutrition’ foods (F(1,4) = 14.73, P < 0.01) and composite
preferences for ‘low nutrition’ foods (F(1,4) = 10.06, P < 0.05). Both
experimental and control groups expressed stronger preferences for
‘pro-nutrition’ foods at post-test. Repeated measures ANOVA and
analyses of covariance were performed on the consumption score for
each individual food. No significant treatment-by-trials interactions were
obtained on any of the analyses. There was a tendency for experimental
group children to consume more of the ‘pro-nutrition’ foods at post-test
than children in the control group, but the differences were not
significant. Overall the pro-nutrition tape increased knowledge but did
not change preferences or consumption.
Radkar
&
Mundlay, 2001
India
Cross-sectional
survey
759 children and 788
adults from the state of
Maharashtra,
India,
from both urban and
rural areas. 60% of the
children were aged
between 10 and 14
years
When asked to list, unprompted, three recalled adverts, the adverts
most frequently recalled by children were for noodles, biscuits, soft
drinks and chocolates. Children showed higher levels of advertising
recall than adults for all the four products except noodles. Parents
reported ‘child’s demand’ for the product was a substantial influence on
buying decisions for several categories of food product.
No
Reeves & Atkin,
1979
Usa
Observational
study
100 mother–child pairs
were observed and
interviewed. Children
ranged in age from 3 to
13 years (mean 7.53
years), and were of
both sexes (46% male,
54% female).
The
socioeconomic status
of the families was
Children initiated 58% of the cereal and confectionery purchase
interactions. In 32% of the interactions the child demanded a particular
product (as opposed to requesting it). Just under two-fifths of
interactions were initiated by the mother, either inviting the child to
select a brand (18%), directing the child to select the brand chosen by
the mother (16%), or choosing the brand without discussion with the
child (4%).
Yes
Mothers agreed to 55% of children’s requests and demands for cereal or
confectionery products, refused 21% of requests and demands, diverted
A-105
slightly above average
11% with suggestions for an alternative product, and ignored the
remainder. Where mothers invited children to select a brand, 70% of
children did so, and the remainder did not make a selection. Where
mothers directed that a brand was chosen, children nearly always
agreed, with only two refusing the selection or trying to divert the mother
to an alternative brand. Conflict over the choice of product occurred in
14% of interactions, with the approximate mean length of argument
being 16 seconds. Ten percent of interactions involved raised voices
and 7% involved ‘verbal aggression’.
Children requested a mean number of 1.6 products, while the actual
number of products purchased was a mean of 1.5, indicating that the
same number of products were purchased as were requested. Thirtyeight percent of children gave reasons for selecting a particular product.
The most common reason was that the product ‘tasted good’ or ‘I like it’
(40% of expressed reasons). Other reasons, receiving only a small
number of responses, included a sweet taste, nutritional value,
television adverts, or free gifts.
Thirty-eight percent of children had watched Saturday morning television
before coming to the store. There was no relationship between whether
children reported watching of television and probability of making a
request or demand for a product. However, there was a significant
relationship between amount of Saturday morning viewing and
frequency of requests and demands, with children who had watched
more Saturday morning television before the trip making more demands
or requests for products than children who had watched less Saturday
morning television (t = 1.69, d.f. = 36, P < 0.05). Other television
exposure measures showed no significant relationship with frequency of
requests or demands while shopping, although there was a tendency for
children with higher television exposure to initiate requests and
demands more frequently.
Mothers perceived that cereal and confectionery adverts had a strong
(50%) or moderate (46%) impact on their children. Sixty-four percent of
mothers said that their children paid close attention to Saturday morning
advertising and 30% that their children paid some attention. The majority
(61%) of mothers described themselves as yielding to some of children’s
requests for cereals and confectionery, while 27% said they did not yield
very often; 12% said they yielded ‘most of the time’. Frequency of
reported yielding to children’s requests was significantly related to
higher levels of children’s Saturday morning television viewing as
reported by mothers (r = 0.27, P < 0.05). Yielding to requests was
A-106
significantly related to having more lenient rules about eating sweets.
Mothers who perceived that advertising had a strong influence on their
children were also significantly more likely to say that their children paid
close attention to advertising (r = 0.45, P < 0.05).
Riecken
Yavas, 1990
&
USA
Cross-sectional
survey
152 children aged 8–
12 years attending a
‘university-operated
school in a [USA] midwestern city
Children’s attitudes towards advertising in general and towards
advertising in the three product categories differed, with children having
more favourable attitudes towards toy adverts than to adverts in the
other product categories or adverts in general.
No
The relationship between attitudes towards advertising in the three
product categories and evaluations of the eight brands was examined to
assess whether children’s evaluations of brands are influenced by their
pre-existing attitudes towards adverts. An association was found for only
one of the three cereal brands and one of the toy brands, suggesting
little clear relationship between attitudes to advertising and brand
evaluations.
Ritchey & Olson,
1983
USA
Cross-sectional
survey designed to
measure
family
factors which might
moderate
preferences
for
sweets in children
122 pre-school children
of both sexes, ranging
in age from 36 to 64
months (mean age
52.2 months), and their
parents
t-tests and analysis of variance revealed few consistent relationships
between family characteristics and parental and child attitudes and
behaviours. Amount of television watched was related to the greatest
number of other variables (the data are not presented in the article), and
was therefore entered into multiple regression analyses.
When the dependent variable was children’s self-reported preference for
sweet foods, none of the independent variables had a significant
relationship with preferences. When children’s frequency of
consumption of sweet foods as reported by parents was the dependent
variable, three independent variables had a significant relationship with
consumption: parents’ own frequency of consumption, amount of
television watching, and parents’ attitudes towards sweet foods.
Together these variables accounted for 35% of the variance in children’s
frequency of consumption. Television watching made a significant
contribution at the P < 0.01 level independently of other variables in the
model. The analyses also examined whether the strength of the
relationships between the variables changed depending on whether the
preschool child was the oldest in the family or not. Associations were
found to be stronger when the preschool child was the oldest in the
family: parents’ own frequency of consumption, amount of television
watching, and parents’ attitudes towards sweet foods together
accounted for 54% of the variance in children’s frequency of
consumption when this child was the oldest, compared with only 12%
when there were other older children in the family. Television watching
A-107
Yes
made a significant contribution at the P < 0.01 level, only for those
children who were the oldest in the family, independently of other
variables in the model.
Robinson et al.,
2007
USA
Experimental study
63
3-5
year
old
children
(mean±SD
age, 4.6±0.5 years;
range, 3.5-5.4 years)
from preschools for
low-income children in
from 6 centers in San
Mateo
County,
California.
By the early age of 3 to 5 years, low-income preschool children
preferred the tastes of foods and drinks if they thought they were from
McDonald’s, (the mean±SD total taste preference score across all food
comparisons was 0.37±0.45 (median, 0.20; interquartile range, 0.000.80) and significantly greater than zero (P_.001)), demonstrating that
brand identity can influence young children’s taste perceptions. This
was true even for carrots, a food that was not marketed by or available
from McDonald’s. These taste preferences emerged despite the fact that
3 of the foods were from McDonald’s and only the branding was
changed, indicating that the effects were not due to familiarity with the
taste or smell of McDonald’s food. Even the children with the lowest
frequency of eating food from McDonald’s had average positive total
preference scores, indicating they preferred more of the branded foods.
Yes
Moderator analysis found significantly greater effects of branding among
children with more television sets in their homes and children who ate
food from McDonald’s more often.
Ross et al. 1980;
1981
USA
Randomized
controlled
experiment
100 children (52 boys,
48 girls) from USA
school
grades
kindergarten to six
At baseline, all respondents were less accurate in their rating of artificial
fruit products than in their ratings of fruit and non-fruit products.
Significant main effects were found for age (F(1,88) = 13.22, P < 0.001)
and for fruit content (F(2,176) = 91.09, P < 0.001). Older children’s
ratings of artificial fruit products were more accurate than younger
children’s ratings, but less accurate than their own ratings for fruit and
non-fruit products.
It was hypothesized that exposure to the experimental adverts would
lower children’s accuracy ratings for the artificial fruit products but would
improve or not affect their accuracy ratings for fruit and non-fruit
products, i.e. there would be a three-way interaction of treatment x
session x fruit content, and that age would enter the interaction if the
adverts had differential effects on younger and older children. No
significant effects on children’s accuracy ratings were found following
naturalistic exposure to the adverts (i.e. adverts embedded in a
television programme). However, planned comparisons computing
interactions of treatment and session indicated that following intensive
exposure to the adverts, accuracy ratings for artificial fruit products
decreased in children exposed to the adverts compared with control
group children (F(1,87) = 5.97, P < 0.05). Accuracy ratings were
A-108
Yes
significantly different between experimental and control group children
following intensive exposure, indicating that experimental group children
became less accurate and control group children more accurate (F(1,87)
= 6.64, P < 0.05).
It was also hypothesized that accuracy ratings by experimental group of
children for artificial fruit products would decrease more for those
products which had appeared in the six adverts than for the six products
which were not advertised (while accuracy ratings for fruit and non-fruit
products which were advertised were expected to increase or remain
the same in comparison with non-advertised products). A three-way
interaction was predicted between session x fruit content x
advertisement. No significant effects on accuracy ratings for artificial or
non-fruit products comparing advertised and non-advertised products
were found after naturalistic exposure to the adverts (i.e. with television
programmes). Planned comparisons computing interactions of treatment
and session indicated that for real fruit products, children became
slightly more accurate after seeing the adverts for the six advertised
products, and slightly less accurate for the non-advertised products
(F(1,47) = 6.26, P < 0.05). This suggested a positive effect of advertising
on accuracy ratings about real fruit products, but the results were
complicated by baseline scores for non-advertised real fruit products
being higher than baseline scores for advertised fruit products.
Following intensive exposure to the adverts, there was a significant
interaction as predicted between session, fruit content and
advertisement. There was a significant difference between accuracy
ratings for advertised artificial fruit products following intensive exposure
and accuracy ratings for non-advertised artificial fruit products (F(1,47) =
9.26, P < 0.01).
Effects of repeated testing were also examined for different age groups.
Older respondents became more ‘sceptical’ with repeat testing (i.e.
more likely to say that products did not contain fruit), while respondents
in the middle age group became more likely to say that any product did
contain fruit. There was no consistent pattern in response for the
youngest respondents.
The investigators concluded that the data indicate a consistent
misjudgement by children of whether artificial fruit products contain fruit.
This misjudgement occurred for all respondents, at all sessions
(including baseline), and in relation to both advertised and nonadvertised products. Viewing of adverts in an intensive situation (i.e.
with no accompanying television programmes) increased rather than
A-109
decreased this tendency to misjudge. This tendency occurred when
comparing: (a) the experimental group’s ratings with their baseline
ratings; (b) the experimental group’s ratings for the advertised products
with their ratings for the non-advertised products; or (c) the experimental
group’s ratings with the control group’s ratings for the same products.
This indicates a robust effect attributable to the adverts. Viewing the
adverts in a naturalistic situation neither improved nor worsened
children’s tendency to misjudge the fruit content of artificial fruit
products.
Sjölin, 2008
Sweden
Review
Literature and reports
with European data
In 2003, the report by Technical Experts (TRS 916) showed probable
evidence between heavy marketing of energy-dense foods and fast food
outlets and an increased risk for weight gain and obesity. Five major
reviews of the evidence on the impact of food marketing to children
(published 1987-2006) showed that marketing inter alia has an
impact on attitudes, purchase requests and consumption.
Nevertheless the International Chamber of Commerce maintains that
there is no scientific evidence that restrictions on advertising would have
an impact on the incidence of obesity and that advertising bans would
be disproportionate and ineffective.
No
Studies have documented that a high percentage of advertisements
targeting children feature sweets, fast foods, and snacks and that
exposure to such advertising increases consumption of these products.
The problem is that the great majority of foods that are advertised are
high in fat, salt and/or sugar (CFC, 2007).
Concludes that marketing of unhealthy foods is thus a contributing factor
to overweight and obesity.
Stoneman
Brody, 1981
&
USA
Randomized
controlled
experiment
th
80 4 grade children
attending
a
rural
elementary school
Factorial analysis of variance (4 x 2) indicated that there was a
significant main effect for experimental condition on number of salty
snacks selected. Newman-Keuls post hoc comparisons were utilized to
define these significant effects further. Children in Condition 1 (adverts
only) selected salty snacks more frequently than children in the control
condition (mean scores 5.35 vs 3.75, P < 0.01). Children in Condition 2
(adverts plus peer modelling similar food choices) selected salty snacks
more frequently than children in Condition 1 (mean scores 6.8 vs 5.35, P
< 0.01). Children in Condition 3 (advert plus peer modelling dissimilar
food choices) selected salty snacks less frequently than children in
Condition 1 (mean scores 3.8 vs 5.35, P < 0.01).
Analysis of variance also revealed a significant interaction between
A-110
Yes
experimental condition and race of child. White children in Condition 3
(advert plus peer modelling dissimilar food choices) selected salty
snacks more frequently than white children in the control condition (4.8
vs 3.1, P < 0.05), whereas black children in Condition 3 selected salty
snacks less frequently than black children in the control condition (2.7 vs
4.4, P < 0.05). Retention scores, both of advertised foods and of peer
model’s choices, were extremely high for all subjects.
Overall the study showed that the adverts alone influenced snack
selection by increasing the frequency with which salty snacks were
chosen. It also showed that the peer modelling influenced snack
selection over and above the adverts, but the direction of this influence
depended on whether or not the peer models made similar or dissimilar
food choices. Furthermore, there was a significant interaction between
experimental condition and race of child that showed white and black
children reacting differently to peers modelling dissimilar food choices.
Black children exposed to the adverts and peer dissimilar modelling
chose salty snacks less frequently than black children in the control,
while white children exposed to the adverts and peer dissimilar
modelling chose salty snacks more frequently than white children in the
control.
Stoneman
Brody, 1982
Tantivejakul,
2001
&
USA
Randomized
controlled
experiment
Thailand
Focus
interviews
group
36 mothers and their
preschool
children
(aged
3–5
years)
recruited from a child
and family centre in
Georgia, USA
Children exposed to the experimental tape engaged in more PIAs in
general than children exposed to the control tape (X = 43.20 vs X =
28.36, P < 0.01), and engaged in more PIAs for the specific products
advertised on the experimental tape (X = 4.4 vs X = 1.9, P < 0.025). The
frequencies of parental “no’s”, verbal put backs and physical put backs
were summed to create a measure of parental power assertion. Mothers
of children exposed to the experimental tape used power assertion
responses more frequently than mothers of children exposed to the
control tape (X = 4.7 vs X = 2.7, P < 0.05). They also made more
alternative offers in response to children’s purchase requests (X = 2.1 vs
X = 2.1 [sic], P < 0.025). No significant difference was found between
experimental group and control group children in number of hours of
television reportedly viewed per week, and the two groups of children
did not differ in the amount of attention they paid to the experimental
tape, which increases the likelihood of the observed differences in
behaviour being attributable to the experimental tape.
Yes
20 focus groups were
undertaken with 120
teenagers (plus 4 initial
pre-test groups)
Television adverts generated cultivation effects and these were different
among heavy and light television viewers.
No
Heavy TV users were found to rely more upon television adverts as a
A-111
product information source, demonstrated by their heavy referral to
advertising during discussions. Heavy users also admitted that when
watching the appeal stuck in their mind and they promised themselves
to buy the product. However this effect was only shown for product
categories specifically targeted at teenagers (such as shampoo and
conditioners and skin care products) and among those with their own
purchasing power.
Heavy TV users were also found to consistently use language very
similar to the form used in advertising. Some heavy TV viewers
expressed similar thinking to themes/portrayals used in advertising.
Overall conclusions: heavy TV viewers tended to use advertising as
their main source for product information, and advertising appeals were
also shown to attract the attention of heavy TV viewers and persuade
them to buy. This effect was not shown for the only food and drinks
product category (milk and dairy products).
(N.B. results don not really separate out the effects for the different
product categories).
Taras
1989
et
al.,
USA
Cross-sectional
survey
66 mothers of children
aged 3–8 years in
public ‘preschools’ and
elementary schools
The most frequently requested food items influenced by television were
sugared cereals (65 requests), sugared fruit (15), fast foods (16), soft
drinks (9) and non-sugared cereals (8). Of the sports items requested,
bicycles were the most popular (12 requests), followed by skateboard
(10), pogo balls (7) and balls (6). High sugar foods made up 66% of
foods requested, followed by high fat items 36%, high salt items 19%,
and low sugar/fat/salt items 7%. High fat items made up 58% of the
items which mothers reported buying in response to requests, followed
by high sugar items 34%, high salt items 22%, and low sugar/fat/salt
items 11%. The authors describe a “strong agreement” between the
relative proportion of foods requested and purchased in the different
nutritional categories (high sugar, high fat, high salt, and low) and the
foods advertised on television, but no statistical significance is reported.
There was a correlation between purchase of requested food items and
purchase of requested sports items (r = 0.44, P < 0.001). There was no
correlation between purchase of sports or food items with child’s
participation in a sport or other physical activity after a televisioninfluenced request.
Significant positive correlations were found between hours of television
viewing and number of food items requested as influenced by television
(r = 0.31, P = 0.006), number of food items subsequently purchased (r =
A-112
Yes
0.44, P = 0.001) and caloric intake as measured by the food frequency
questionnaire (r = 0.34, P = 0.001). Snacking while watching television
was also significantly positively correlated with number of food items
requested and purchased and with caloric intake. Watching television
during a meal was significantly negatively correlated with caloric intake.
Correlations were also examined between food requests and purchases
influenced by television and nutrient intake as assessed by the food
frequency questionnaire. Total food requests and purchases were
significantly correlated with saturated fat and sugar consumption (P =
0.012 and P = 0.001 respectively), but not with salt intake. Requests for
and purchase of high fat foods were significantly correlated with
saturated fat intake (
P = 0.012), sugar intake (P = 0.001) and salt intake (P = 0.004).
Requests for and purchases of high sugar foods were significantly
correlated only with sugar intake (P = 0.03), while requests for and
purchases of high salt foods were not correlated with salt intake.
Relationships between hours watching television and requests and
purchases of sports items and physical activities influenced by television
were generally weaker. There was a significant negative association
between hours watching television and sport items purchased (P =
0.009).
Taras
2000
et
al.,
USA
Cross-sectional
survey
237 American families
The correlation coefficients between (i) the scores of advertised product
categories and scores of products requested, and (ii) the scores of
advertised product categories and products purchased were significant
at 0.91 (P < 0.0001; t = 21.08) and 0.94 (P < 0.0001; t = 16.92)
respectively.
No
To account for the predominance of two highly correlated categories
(restaurants and sugared meals), an additional analysis was undertaken
using only the other 15 product categories. The correlation coefficients
were still significant at 0.61 (P < 0.015) and 0.66 (P < 0.008),
respectively.
Unnikrishnan
Bajpal, 1996
&
India
Cross-sectional
survey
and
qualitative
interviews
with
focus groups
730 children aged 5–
15 years in Delhi
Pepsi adverts were the favourite cold drink adverts among the sample
as a whole, and among upper class and middle class children. Pepsi
adverts were second favourite among lower class children, who tended
to prefer adverts for cheaper Indian brands. Pepsi was also the most
frequently consumed cold drink. Children’s preferred cold drinks and
consumption habits were also reported, and the authors claimed that
preferences and consumption habits reflected advertising patterns,
A-113
No
although no statistical analyses were conducted to investigate the
relationships.
Utter, Scragg &
Schaaf, 2006
New Zealand
Secondary
data
analysis
of
a
nationally
representative,
cross-sectional
survey
3275 children aged 5
to 14 years surveyed in
homes or schools of
New Zealand
The odds of being overweight or obese increased with duration of TV
viewing for children and adolescents when controlling for age, sex,
ethnicity, socio-economic status and physical activity. Time spent
watching TV was an independent correlate of obesity, when controlling
for age, sex, ethnicity, SES and physical activity, among children (odds
ration 2.1; 95% confidence interval (CI)1.1-4.0) and adolescents (OR
2.9; 95% CI 1.5-5.7).
No
Children and adolescents who watched the most TV were significantly
more likely to be higher consumers of foods most commonly advertised
on TV: soft drinks and fruit drinks, some sweets and snacks, and some
fast foods. Both children and adolescents watching two or more hours of
TV a day were more than twice as likely to drink soft drinks five times a
week or more (P = 0.03 and P = 0.04, respectively), eat hamburgers at
least once a week (both P = 0.02), and eat French fries at least once a
week (both P < 0.01). TV use was positively associated with the
consumption of soft drinks, fruit drinks, potato chips, chocolate sweets,
biscuits, hamburgers and French fries (not fried chicken), and negatively
associated with the consumption of fruits and vegetables.
Vajpeyi, 2001
India
Cross-sectional
survey
159 parents and 244
children ‘in and around
Delhi’
Findings relevant to the review:
Exposure to food advertising: Low income consumers in developing
countries are described as being less exposed to television advertising
for fast-foods (e.g. because they have no television) and cannot really
afford it and are therefore ‘largely protected from the fast-food menace
by their so-called poverty.’
Children’s awareness and liking of food advertising: Most children like
adverts for either of the big two Coke brands (Coca-Cola and Pepsi) and
are aware of the ‘war’ in advertising between the two (and the use of
celebrities to endorse and promote products).
Food preferences: Products like chocolate, ‘candies’ and gum are
reported as the most popular among children and teenagers. The author
comments that Indian children ‘buy into imported foods in a big way’.
Parent’s views on food advertising to children: Almost all parents felt
that there was a ‘huge and growing’ influence of food advertising on
children. ‘Most’ parents felt there was a need to restrict fast-food
marketing activity. Parents also said that they felt ‘compelled’ to respond
A-114
No
to children’s demands for food products at times.
Food consumption behaviour: The author comments that ‘fast-food
chains have made their presence felt in Delhi.’ Fast-food was found to
be very popular among children although consumption was usually
restricted to a few days per month. Most children were aware of the lack
of nutritive value in fast-foods and parents revealed their concerns about
the increase in consumption of fast-food and confectionary products
among children. As an aside, it is noted that fast-food products are often
consumed by higher social class Indians, and that ‘low income children
and parents….do not consume fast foods at all.’
Ward, Reale &
Levinson, 1972
USA
Structured
interviews
67 children aged 5–12
years
Favourite advert: The largest category of favourite adverts was food
adverts (33%), followed by toys (10%), programme announcements
(9%), soft drinks (6%), cars (5%), and PSAs (3%). Seventeen percent of
favourite adverts were for other products. One-fifth of children did not
have a favourite or could not remember it. Food adverts were also the
most disliked adverts: 15% of least favourite adverts were for food
products. In over a third of cases, children could not remember their
least favourite advert or did not have one.
No
Other results do not relate to food advertising.
Wiecha
2006
et
al.,
USA
Prospective
observational study
Williams, 1974
USA
Cross-sectional
survey
Wiman
&
Newman, 1989
USA
Cross-sectional
survey
Five public schools in 4
communities
near
Boston. The sample
included 548 students
(mean age at baseline,
11.70 years; 48.4%
female; and 63.5%
white). baseline (fall
1995) and follow-up
(spring
1997)
measures of youth diet,
physical activity, and
television viewing
54 9–13-year-olds
Main outcome measures were change in total energy intake and intake
of foods commonly advertised on television from baseline to follow-up.
After adjusting for baseline covariates, each hour increase in television
viewing was associated with an additional 167 kcal/d (95% confidence
interval, 136-198 kcal/d; P<.001) and with increases in the consumption
of foods commonly advertised on television. Including changes in
intakes of these foods in regression models provided evidence of their
mediating role, diminishing or rendering nonsignificant the associations
between change in television viewing and change in total energy intake.
Among youth increases television viewing predict increases in total
energy intake and that consumption of foods commonly advertised on
television mediates this relationship.
No
All children reported spending “almost half” of their allowance each
week on snacks, and 44% reported that they purchased snacks that
they saw advertised on television.
No
327 children in USA
elementary
school
grades 3–6 (ages 8–12
One-way analysis of variance was conducted to investigate whether
television viewing varied significantly by age. This revealed that there
was no variation by age in afternoon viewing, but that younger age was
Yes
A-115
years)
significantly related to greater viewing of Saturday morning television (P
< 0.000), while older age was significantly related to greater viewing of
weekday evening television. Age correlated significantly with nutritional
knowledge and understanding of nutritional phraseology, with older
children having higher scores on both measures.
Partial correlation coefficients were calculated for the relationship
between television viewing and the two nutritional measures controlling
for age. The amount of television viewed on Saturday mornings
correlated negatively with nutritional knowledge (r = −0.116, P < 0.05)
and understanding of nutritional phraseology (r = −0.113, P < 0.05) i.e.
greater exposure to child-oriented television was associated with less
nutritional knowledge and understanding. There was no significant
relationship between television viewing on weekday afternoons and
either of the nutritional measures. Weekday evening (i.e. non-childoriented) viewing was positively correlated with nutritional knowledge (r
= 0.114, P < 0.05), i.e. the more television viewing the greater
knowledge. There was no significant relationship between academic
achievement (as measured by academic grade level) and amount of
television viewing at any of the time periods examined. There was
however a positive relationship between academic grade and nutritional
knowledge (r = 0.304, P < 0.001) and understanding of nutritional
phraseology (r = 0.297, P < 0.001).
Overall, when controlling for age, poorer nutritional knowledge and
understanding were associated with greater viewing of television at
times when advertising is more child-oriented. The possibility that poorer
nutritional knowledge/understanding and greater television viewing
could both have been associated with, or caused by, some other factor
is not ruled out, as the results showed that there was a stronger
relationship between academic grade and nutritional knowledge scores
than between Saturday television viewing and nutritional knowledge.
However, Saturday morning television viewing did not vary significantly
by academic grade, suggesting that television viewing or some other
variable accounted for some of the relationship with nutrition knowledge.
Wong
1992
et
al.,
USA
Cross-sectional
survey
1081 children aged 2–
20 years (mean age
7.4 ± 3.6 SD) in
California
Children with higher cholesterol levels were more likely to have a parent
or grandparent with high cholesterol (P = 0.02), to consume lean meat
(P = 0.01), to have fat trimmed from meat (P = 0.02), to have food
cooked in vegetable oil (P = 0.04), and to watch two or more hours of
television/video per day (P = 0.001). The use of television watching as a
predictor variable together with family history predictors identified 85%
of the children with higher cholesterol levels. Only 66% of this group
A-116
Yes
would have been identified without the use of television watching as a
predictor.
Children who reported watching more than four hours of television daily
were less likely to consume lean meat (P = 0.006) or engage in physical
activity (P = 0.02).
Multiple logistic regression analyses with high cholesterol in children as
the dependent variable found that family history of high cholesterol,
higher levels of television viewing, and lean meat consumption were
each independently associated with increased risk of high cholesterol.
Children watching 2–4 hours of television daily were approximately twice
as likely (relative risk 2.2, P < 0.01), and those watching ≥ 4 hours four
times as likely (relative risk 4.8, P < 0.01), to have a high cholesterol
level than children watching < 2 hours daily. The relative risk for family
history of high cholesterol was 1.6 (P < 0.05), and that for lean meat
consumption 2.5 (P < 0.01).
Yavas & AbdulGader, 1993
Saudi Arabia
Cross-sectional
survey
217 students in Saudi
school grades 5–8;
56% of the sample
were male and 44%
female
Two-fifths of respondents reported watching three or more commercial
breaks per day (the highest response on the scale). There were no
significant gender differences. The most frequently recalled adverts
were for foods (ranked first), followed by soft drinks, toys and cars.
Spearman’s rank order correlation indicated moderate agreement in the
rankings by males and females, with the order for males being food, soft
drinks, cars and clothes; for females, food, soft drinks, toys and baby
care products. The most preferred type of advertising was humorous
adverts, liked by 72.3%, followed by cartoons (61.7%) and educational
adverts (39.6%). The most popular types of advert by product category
were food adverts (52.5%), followed by cars (47.9%), soft drinks (40%)
and detergents (39.6%). There were significant gender differences, with
75% of females liking food adverts compared with 34.7% of males (P <
0.05); females also liked soft drinks adverts more than males (50% vs
32.2%, P < 0.05).
Around a third of respondents said that they ‘always’ and 46%
‘sometimes’ asked parents to buy items they had seen advertised, and
that parents agreed ‘always’ (43.9%) and ‘sometimes’ (45.3%). A
majority (69.3%) said that they could always identify advertised brands
in the shops.
A-117
No
ISBN 978 92 4 159883 5
A-118

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