Two of the institutional actors that have been carefully considering all of the questions above are funders and publishers. They need to take on these issues because they must develop policies for applicants and grantees, and authors, to follow. Given the considerations, benefits, and challenges above, how should key institutional actors design policies to encourage responsible data sharing? Pretend you are the editor-in-chief of a journal of your choice. Develop a draft policy that you will translate into author guidelines for providing the qualitative data that underpin articles published in your journal, and accompanying “materials” (e.g., documentation). Your policy should address the following issues:
Which data should be shared and what in addition to data needs to be shared (“materials”)?
When in the publication process should data (and materials) be shared?
Where should data (and materials) be shared?
What established exceptions should there be?
Who should judge whether a scholar’s situation fits within those exceptions – or, if it does not, whether an exception should nonetheless be made?
There are no right or wrong answers to these questions! We encourage you to evaluate your answers along the following lines:
Is your policy coherent – do its different aspects fit together without contradicting each other?
How easy will it be to translate your policy into clear guidelines for authors?
What kind of behavior are the different aspects of your policy, and your policy as a whole, encouraging on the part of authors and reviewers?
Of the various involved actors (authors, reviewers, other institutions), whom have you empowered – granted decision-making authority?
Have you addressed possible trade-offs between promoting data sharing and deterring authors from submitting their best work to your journal?
What potential sources of conflict does your policy introduce?
To illustrate the state of current practices, here are two different data policies. The PLOS journals data policy mandates data deposit for all articles, explicitly states that proprietary data are not acceptable as the sole empirical basis for an article, and describes specific options for qualitative data. The American Political Science Review’s data policy (see “Reproduction Policy”) is significantly shorter with fewer requirements. As you consider these policies, note some of the differences between what is required for scholarship based on quantitative data and analysis v. work based on qualitative data and analysis.