"How Many Scientists Fabricate and Falsify Research?"

"A Systematic Review and Meta-Analysis of Survey Data"
Daniele Fanelli, PLoS One, May 29, 2009.

From the Abstract:

A pooled weighted average of 1.97% (N = 7, 95%CI: 0.86–4.45) of scientists admitted to have fabricated, falsified or modified data or results at least once –a serious form of misconduct by any standard– and up to 33.7% admitted other questionable research practices. In surveys asking about the behaviour of colleagues, admission rates were 14.12% (N = 12, 95% CI: 9.91–19.72) for falsification, and up to 72% for other questionable research practices. Meta-regression showed that self reports surveys, surveys using the words “falsification” or “fabrication”, and mailed surveys yielded lower percentages of misconduct. When these factors were controlled for, misconduct was reported more frequently by medical/pharmacological researchers than others.

Considering that these surveys ask sensitive questions and have other limitations, it appears likely that this is a conservative estimate of the true prevalence of scientific misconduct.

Comment: I would agree with Fanelli that the self reported occasions of misconduct probably under estimate the true value.

Lets think a moment about the estimate of 14% of scientists who report knowledge of falsification by at least one colleague and 72% who report knowledge of other questionable practices by at least one colleague. How do people define "colleague" in answering such a question? I would think that all scientists have read about some of the widely publicized cases of falsification by scientists, so presumably they are not defining "colleague" broadly enough to include any scientist or even any scientist working in their own field of science.

How many "colleagues" would a scientist have on the average with whom he/she was working sufficiently closely to directly observe misconduct? I would guess perhaps four or five at a given time, with some scientists who do a great deal of collaborative research at an extreme tail of the distribution. On the average, perhaps surprisingly, scientists don't spend as much as a decade doing research, so the total number of other scientists observed would be relatively modest. And of course, one does not closely observe the conduct of friends and colleagues.

I recall one fairly strong argument during my days as a researcher in which my coauthors and I debated (hotly) how to report our study, a debate that resulted in doing what I felt to be the right thing. How would one report the attitude of the colleagues seeking to present the results "in the most favorable light"?

And of course, I would be reluctant to answer such a question positively in any case. It would involve a charge against a colleague, which even in the case of anonymity is distasteful.

The level of questionable conduct may be quite high. JAD