UWM Data Publishing Guide

UWM Data Publishing Guide
UWM Data Publishing Guide
Researchers with NSF, NIH1, DOE, and other federal funding are now expected to make data from their
research publicly available. Additionally, journals such as PLOS, Science, Nature, BMJ, and more require
researchers to make their data available after publication and/or during peer review. This guide is
intended to help UWM researchers navigate these new requirements for data sharing.
If you have further questions about data management and sharing, contact [email protected]
You can also find more information at:
UWM Data Services: http://uwm.edu/libraries/dataservices/
Data management guide: http://guides.library.uwm.edu/data
When to share
Data sharing happens once you publish. After you publish an article, white paper, etc. you should
make the data that support that research publicly available.
What to share
Share data that support your publications and anything necessary to reproduce your results. The
form of data to share (raw, cleaned up, analyzed, etc.) varies by field and often depends other
considerations, like file size. Use your best judgment as to what is the most useful form for your
data.
Sensitive data, such as that containing personally identifiable information, should not be shared.
However, do consider sharing if you can properly anonymize or redact the data. Corporatesponsored research may also have similar restrictions on sharing; contact Mark Doremus
([email protected]) if you have private funding and want to share your data.
Why to share
Data sharing has been shown to increase citations for the corresponding article (see
https://peerj.com/articles/175/). Data sharing also makes it easier for you to find and reuse data
later as the data live in a stable location. Finally, all major federal funding agencies will soon start
requiring data sharing if they do not already (see http://1.usa.gov/1jg1QXt).
Preparing to share
Perform quality control on your data prior to sharing to clean up errors and inconsistencies.
OpenRefine (http://openrefine.org/) is particularly recommended for cleaning up tabular data.
1
NIH grants for over $500,000 per year are required to make the data publicly accessible
October 2014
UWM Data Publishing Guide 1/2
Use non-proprietary or common file types (.csv, .txt, etc.) whenever possible. This makes data files
easier to open and more long lived.
Shared data should also be documented. Use a README.txt file, a data dictionary, or some other
form of documentation so that others can understand your data. Data Services’ data management
guide has more information on documentation formats: http://guides.library.uwm.edu/data.
It is also recommended to license your data with a Creative Commons license
(http://creativecommons.org/), preferably CC0 as you cannot copyright most research data.
Copyright may apply to some types of data, such as images, but the majority of data exist outside of
copyright. An open license clears up any confusion over copyright and what rights are available for
the data.
For more recommendations, see the article “Nine simple ways to make it easier to (re)use your
data”: http://library.queensu.ca/ojs/index.php/IEE/article/view/4608
Where to share
Share your data in a data repository. This is preferable to sharing on a personal website or by
request because repositories are more stable and this method of sharing requires no work from you
after deposit. Additionally, many repositories have built-in methods for citing your data.
To find a repository for your data, consider using one of the following:




figshare (general): http://figshare.com/
Dryad (biology): http://datadryad.org/
ICPSR (social science): https://www.icpsr.umich.edu
UWM Digital Commons (small, discrete datasets): http://dc.uwm.edu/
There are many other general and disciplinary repositories available. Contact Data Services at [email protected] for a recommendation.
Other considerations
Consider sharing your research code via Google Code (https://code.google.com/) or GitHub
(https://github.com/), which helps with reproducibility.
Add datasets to your CV and report them as products of your grant – this is especially recommended
when applying for your next grant. You can also sometimes track metrics (views, downloads, etc.)
via the data repository or ImpactStory (https://impactstory.org/).
October 2014
UWM Data Publishing Guide 2/2
Was this manual useful for you? yes no
Thank you for your participation!

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

Download PDF

advertisement