Journal data policies: Exploring how the understanding of editors and authors corresponds to the policies themselves.

Despite the increase in the number of journals issuing data policies requiring authors to make data underlying reporting findings publicly available, authors do not always do so, and when they do, the data do not always meet standards of quality that allow others to verify or extend published results. This phenomenon suggests the need to consider the effectiveness of journal data policies to present and articulate transparency requirements, and how well they facilitate (or hinder) authors' ability to produce and provide access to data, code, and associated materials that meet quality standards for computational reproducibility. This article describes the results of a research study that examined the ability of journal-based data policies to: 1) effectively communicate transparency requirements to authors, and 2) enable authors to successfully meet policy requirements. To do this, we conducted a mixed-methods study that examined individual data policies alongside editors' and authors' interpretation of policy requirements to answer the following research questions. Survey responses from authors and editors along with results from a content analysis of data policies found discrepancies among editors' assertion of data policy requirements, authors' understanding of policy requirements, and the requirements stated in the policy language as written. We offer explanations for these discrepancies and offer recommendations for improving authors' understanding of policies and increasing the likelihood of policy compliance.
Competing Interests: I have read the journal's policy and the authors of this manuscript have the following competing interests: TC and AG are employees of the Odum Institute Data Archive. TV was a member of the Dryad Digital Repository Board of Directors and participated in unpaid activities for Dryad during the research project period. EH is an employee of Dryad Digital Repository. This does not alter our adherence to PLOS ONE policies on sharing data and materials.