Reproducibility

Improving the quality of publicly archived data

+ See all authors and affiliations

Science  15 Jan 2016:
Vol. 351, Issue 6270, pp. 237-238
DOI: 10.1126/science.351.6270.237-b

The deposition of research data in public archives and repositories is becoming common practice across the sciences, with many journals mandating the archiving of data associated with papers. However, in a survey of 100 publicly archived ecological and evolutionary data sets associated with recently published papers, Roche et al. found that fewer than 50% were complete and reusable. Many lacked necessary metadata, consisted of processed rather than raw data, or were presented in non-machine-readable formats, all of which are obstacles to reusing published analyses. The situation might be improved through the adoption of a reward culture, in which carrots rather than sticks for the public deposition of fully reusable data become the order of the day.

PLOS Biol. 10.1371/journal.pbio.1002295 (2015).

Related Content

Navigate This Article