The journal Epidemiology has refused to support transparency guidelines. In its editorial it gives the reason: research transparency hinders novelty and creativity.
With a range of scandals questioning the trust we have in science, more and more initiatives try to promote research quality. For example, the Transparency and Openness Promotion (TOP) Guidelines support of the principles of openness and reproducibility. For example, the guidelines state that journals should support data transparency and encourage the citation of data sets (as a reward for authors who publish their data). 510 journals across all disciplines have already subscribed to the standards.
It comes as a surprise that the journal Epidemiology does not see such standards as useful. In contrast, the journal fears that the authors, reviewers, and editors would view papers through the lens of transparency, rather than focusing on “quality of the research and the clarity of its presentation.” Implementing the TOP guidelines would hinder “counter” the journals’ goals of “creativity and novelty.”
The journal also criticizes that transparency guidelines would bring “the increased burdens” on authors, reviewers and editors. Finally, the journal states that the TOP guidelines may soon be outdated, and that such guidelines (which in encourage replication) would falsely lead the research community to make statements about ‘right and wrong’ papers.
A few thoughts:
- Yes, working transparently takes time. But how will the research community be able to trust results without seeing the data and code, or without seeing a pre-registration plan? If journals refuse to support transparency, then I don’t see how we can ever produce legitimate “novel and creative” results.
- Yes, the guidelines may become outdated at some point. But subscribing to them now would support at least minimum standards that the scientific community has agreed on. Who says that there won’t be revisions in the future to adjust the guidelines to new insights into transparency?
- Yes, it does not make sense to refute a paper as being ‘wrong’ or claim that it is ‘right’ just because of one single replication attempt. But that is not the goal of replication studies, and it’s not the goal of transparency initiatives. I felt hat different discussions are mixed up here.
None of the points raised by Epidemiology make sense to me, and I wonder if the journal editors fear more workload for themselves, or a decline in submissions from authors who don’t want to take the time to work transparently? It strikes me as very odd for a journal – in these times of daily scandals about scientific misconduct – to take a position like this.
Here are the TOP guidelines.