Data Sharing

Continuing on last week’s discussion about data, I read a blog post by Cameron Neylon about the reluctance of authors to share their data. The post was in response to a paper titled Empirical Study of Data Sharing by Authors Publishing in PLoS Journals, published last week in PloS ONE.

The paper details a study in which 10 authors of PLoS articles were asked to share their data for use in a Master educational project. PLoS authors were chosen because of the journal’s policy that “publication is conditional upon the agreement of authors to make freely available any materials and information associated with their publication that are reasonably requested by others for the purpose of academic, non-commercial research.”

It stands to reason then, that the authors of articles published by PLoS would be willing to share their data. That was not the case. Only 1 of the 10 authors provided their data. Although the study was small, the results suggest a greater problem. Why are authors so reluctant to share their data even when the publication requires it?

Neylon suggests that the findings of this study are not new, and the reasons the authors gave for not sharing their data are not new either. Some authors did not reply, others were unable to be contacted, some said it was “too hard” or their institution “forbids it.” She goes on to suggest that none of these excuses are acceptable.

Neylon insists that it is time for the publications to step up and take responsibility for ensuring their data sharing policies are followed. She believes that the most effective way to go about it is to pull the papers of authors who refuse to share their data. By making an example of a few authors, the data sharing problems for a particular journal will quickly become a thing of the past as authors choose to either share their information or publish elsewhere. Moreover, the publication will benefit from being seen as a source of quality data, further differentiating it from other journals in the field.

And for those who say sharing their data is “too hard” Neylon states that if an author cannot maintain their data so that it can be used again, they shouldn’t be publishing anyway. Neylon does allow that there are occasions when data must be kept confidential, but she argues that these should be the exceptions- exceptions made only after careful consideration of the individual circumstances.

In theory, requiring publications to follow up on the data sharing practices of their authors sounds like a good idea. But I can’t help but wonder if we are putting too much weight on the publishing industry. Is it really their job to make sure their authors share their data? It seems that the authors should step up and make data sharing happen, but it is also clear that they are not currently doing it. What kinds of fundamental changes need to occur to ensure data sharing?