Biomedical supervisors’ role modeling of open science practices

Abstract

Supervision is one important way to socialize Ph.D. candidates into open and responsible research. We hypothesized that one should be more likely to identify open science practices (here publishing open access and sharing data) in empirical publications that were part of a Ph.D. thesis when the Ph.D. candidates’ supervisors engaged in these practices compared to those whose supervisors did not or less often did. Departing from thesis repositories at four Dutch University Medical centers, we included 211 pairs of supervisors and Ph.D. candidates, resulting in a sample of 2062 publications. We determined open access status using UnpaywallR and Open Data using Oddpub, where we also manually screened publications with potential open data statements. Eighty-three percent of our sample was published openly, and 9% had open data statements. Having a supervisor who published open access more often than the national average was associated with an odds of 1.99 to publish open access. However, this effect became nonsignificant when correcting for institutions. Having a supervisor who shared data was associated with 2.22 (CI:1.19–4.12) times the odds to share data compared to having a supervisor that did not. This odds ratio increased to 4.6 (CI:1.86–11.35) after removing false positives. The prevalence of open data in our sample was comparable to international studies; open access rates were higher. Whilst Ph.D. candidates spearhead initiatives to promote open science, this study adds value by investigating the role of supervisors in promoting open science.

Workflow for detecting biomedical articles with underlying open and restricted-access datasets

“…Third, extraction was performed by two raters in cases of doubt. The high reliability of the outcome is supported not just by the high IRR, but also by two further observations: First, in the meta-research study by Haven et al [ 35 ], manual confirmation of open data status substantially increased the correlation between the open data status ascribed to supervisor and supervisee articles. Second, we observe only a low one-digit number of open data articles which we missed in the screening, according to the post-hoc registration available to Charité researchers.…”

“…In an independent analysis, Haven et al [ 35 ] screened a subset of 2042 articles published by biomedical researchers at four university medical centers (UMCs) in the Netherlands. They found that out of these articles, there were 179 articles with shared datasets detected by ODDPub, and 191 datasets were manually checked in Numbat.…”

“…First, we applied it to publications of researchers at the Charité –Universitätsmedizin Berlin, a large German university medical center with >6000 publications per year. Second, we draw on a separate dataset generated for a meta-research study on the effects of supervision on open science practices [ 35 ]. Here, data were extracted for publications by researchers at four Dutch university medical centers.…”

“…However, this must not only be in the context of institutional incentivization. The results by Haven et al [ 35 ] show a strong effect of manual screening. In their article, they calculated the correlation between supervisor and supervisee data sharing rates.…”

Workflow for detecting biomedical articles with underlying open and restricted-access datasets

“…However, this must not only be in the context of institutional incentivization. The results by Haven et al [35] show a strong effect of manual screening.…”