The final morning of the research integrity meeting began with a question that should probably have come earlier in the meeting: what do we know for certain about bad research behaviour? Is misconduct actually on the rise? Sure, there are more scientists than ever, and competition between them is rising, factors that you think would contribute to more misconduct. There are also worrying signs from the young that internet ´research´ at school and university is becoming a substitute for real academic work. But does that mean this is a problem that is only going to get worse? And what can the science community do about it?
Nick Steneck from the US Office of Research Integrity summarized two decades of US research in this area, and highlighted areas that still need more investigation: what is the harm done to science by questionable research practices? Things like refusal to share data, ghost authorship, misleading citation practices and poorly managed
conflicts of interests. We have anecdotes, but what about the evidence?
One policy question on which some new data was presented at the meeting is whether ethics training and education works. Training in responsible conduct of research is one of the most popular solutions proposed for scientific misbehaviour – and though no-one expects an ethics class to stop a dedicated fraudster, the hope is that it somehow raises the general level of ethical awareness.
The sobering news from Melissa Anderson and her colleagues at the University of Minnestoa is that such training does not work as we might hope. You can access her talk here; the results also appear in the September issue of Academic Medicine.
Melissa analysed data from a 2002 survey of US biomedical scientists who were asked about the amount of formal instruction and informal mentoring in ethics they had received – and how that had affected their subsequent behaviour. There were some positive benefits of instruction – in some cases it improved scientists knowledge of good conduct but didn´t seem to change their behaviour. The bad news is that some types of mentoring actually made things worse. In particular, mentoring by an advisor in research ´survival skills´ actually increased misbehaviour in seven areas that Melissa studied – overwhelming any benefits from formal training.
Melissa thinks we need to train the trainers better. Sometimes the job is left to university compliance officers who have no background in science. Or to online instruction tools that replace proper discussion with a box-ticking exercise. As for mentoring, young researchers still need mentoring in personal, financial, and research ethics – and in the art of survival – but she suggests collective mentoring discussions are a better way to reinforce good behaviour over bad. Scientists are often lousy teachers of ethics, Melissa admits, but she thinks they can do better. Let´s hope so, the future generation of researchers is in their hands.
Leave a Reply