This morning the front page of one of the major newspapers (NRC, see online article in Dutch, but ‘right-click’-> ‘Translate in English’ works pretty well) opened with the following headline:
More than half (51.3%) of the academic researchers responding to a large-scale and nation-wide Dutch survey on research misconduct and questionable research practices (QRPs) indicate frequently engaging in at least one form of QRP (e.g. not publishing/disclosing tests with nonsignificant findings, selectively citing references backing one’s argument while ignoring opposing prior work, not disclosing known flaws or limitations of the research design).
While these numbers are perhaps disconcerting to those outside academia, they are not in the least surprising to me based on prior studies on this matter in other contexts as well as my own (albeit limited and idiosyncratic) experience as a young researcher, and might even be conservative estimates given that participation in the survey was voluntary.
The study further finds perceived publication pressure and junior researcher status to correlate positively, and anticipated stringency of review processes and scrutiny of peers to correlate negatively with engagement in QRPs. While causation is not implied here, for sure such factors corroborate the idea that not only researchers, but also universities, supervisors/senior scholars, as well as journals, editors, and reviewers, all have a role to play in creating an environment promoting responsible and dis-incentivizing/punishing questionable research conduct. Indeed, as noted by Tyge Payne and Duane Ireland a few years ago in an editorial in the flagship journal of my own field, ‘it takes a village’ “[to build]…a stronger (global) community of family business [and other] researchers that supports and promotes ethical conduct in publishing research through increased dialogue and interaction.”
Without having any empirical backing for this notion other than my own observations, I believe it is exactly the currently very limited dialogue and interaction surrounding the ethical aspects of our research practices that may underlie the limited awareness of, and persistent taboo surrounding, how and why we seemingly accept many at minimum questionable practices almost as common conduct.
Again taking my own field as an example, It would for instance really be great to see more attention in journals such as FBR or JFBS, and in workshops or conferences such as IFERA, FERC, or EIASM – especially in doctoral/junior faculty consortia or summer schools – for active, honest, and non-judgmental discussion about the very much human potential to let biases and non-result-neutrality affect our work. In the end, we all want to publish and cite interesting ideas with corroborating evidence, but the long-term value and purposefulness of that will be limited if somewhere in the back of our mind even the slightest doubts exist about the level of care and scientific integrity underlying presented findings.
At least that’s how my mind works.
Here therefore a (most likely non-exhaustive) list of things I am pretty sure I have done in the past years that – had I been more mindful of what I was doing and/or more aware of the ‘why’ such things are undesirable – I may not have done, or at least not so casually.
- Deciding that ‘p < .10’ should also be considered quite significant after observing findings rather than before conducting my analyses.
- Deciding whether or not to transform variables after, rather than before, observing their distributional properties in my particular sample.
- Citing my own paper or those of befriended colleagues even though the added value of doing so is limited and/or there would be better sources available for the same illustrative purpose.
- Not keeping (sufficiently clear) notes about the many micro decisions involved in designing or executing a study.
- Deciding to abandon a project altogether after initial findings were deemed too weak/uninteresting to ever be ‘publishable’ in a reputable journal, without however disclosing these results.
- Overgeneralizing the implications of statistical findings based on a very particular, modestly-sized sample.
How about you? Do you recognize yourself in some of these or other practices that upon further thought might not exactly be examples of responsible research? Feel free to add below, or let me know in private. I would be happy to think about constructive initiatives – in the family business research community or beyond – that would help in promoting awareness and reflective discussion on our research practices.