Collective Openness and Other Recommendations for the Promotion of Research Integrity

Research integrity is essentially a matter of behavior. It is embodied in the actions and decisions of scientists, rather than in the standards, codes, regulations and norms that aim to shape that behavior. Misconduct and other questionable research behaviors stand in sharp contrast to research integrity. Measures intended to promote research integrity should therefore be held to a behavioral standard. If they promote right behavior, they can be judged successful; if they show no association with proper or improper conduct, or if, paradoxically, they show evidence of increasing the likelihood of misconduct by scientists, then they are not successful. This behavioral criterion is both simpler and tougher than more common evaluative criteria. Instruction in the responsible conduct of research (RCR) often relies on assessments of students’ reactions to hypothetical situations, or their understanding of ethical principles, or their knowledge of policies and rules. It is not often judged in terms of its ability to affect scientists’ behavior. Yet, at the very least, we should expect scientists who have been trained in RCR to be less likely to engage in misbehavior of any kind. My colleagues and I have investigated this proposition, and our results are not encouraging. I summarize our empirical findings here and suggest ways to promote research integrity that appear promising in terms of their potential to shape behavior. Without evidence of behavioral impacts, RCR instruction cannot be relied upon to ensure or promote research integrity.