How many scientists fabricate or falsify their research?
Uncategorized — By Dmitry Podolsky on June 3, 2009 at 6:05 pm
Save This Post as PDF
This is the title of a rather intriguing paper recently published in PLos ONE. As it turns out, approximately 72% of researchers have seen at least once how their colleagues used inappropriate or incorrect methods of research.
The study by Dr. Fanelli contains analysis of 28 polls offered for answer to actively working scientists. Among different questions in each poll there were questions like “Did you falsify data in your study at least once?” and “Did you ever notice that your colleagues falsify data of their study or use other questionable practices?”. As a result, about 1.75% of all people entered the poll falsified data in their own work and about 30% used other questionable practices. (Hmm, I wonder what do they mean by questionable practices, teach me one, pleeease, science is so competitive field.)
These numbers differed strongly if the questions were not related with a person who was asked the question, but with his/her colleagues instead. As it turns out, 14% have noticed that their colleagues falsify data in their research and about 72% testified use of other questionable practices in the work of their colleagues (hehe, not surprising at all, is it? Gealous bastards 
).
According to the study, results don’t depend much on country where the poll was conducted, but do depend on the field of research – for example, inappropriate methods and questionable practices are more often used in medicine and pharmacology. Fanelli concludes that scientists who work in medicine and pharmacology are generally more honest than, say, physicists Yes, sure, Dr. Fanelli, and sociologists are even more honest than pharmacologists.
A somewhat similar study was conducted by Washington Post in 2005: people who recently received grants from NIH (National Institutes of Health) were offered to fill a questionnaire. According to the published results of the study, 5% considered possible to neglect facts in the study, 10% were included into a list of authors of a paper they did not contribute to in any way (well, this sounds familiar), 16% changed the results of their study so that they look like more favorable from the point of view of a sponsor of the study. It feels so good that we don’t have many sponsors in theoretical physics…
5 Comments
Let’s hope the physicists are at least as honest as the biomedical scientists: the competition in physics is probably a bit let tough, and there is less money to fight for. Also, the public probably does not expect the physicists to deliver something like the instant cancer cure or the immortality recipe
Yep, money is the key to the problem, I guess. We don’t have sponsors whose sales depend on results of our studies.
I meant “the competition in physics is a bit LESS tough”. Sorry for the typo
Am I reading this correctly? Physicists *are* less honest than medical researchers? I find this hard to believe– since in physics there’s much less room for error when theory compares to data. Maybe some physicists think no one’s going to read their papers anyway? I don’t know. Seems counter intuitive. I’d think the harder the science, the less dishonesty. I’m going to have to read the PLoS article.
Well, the guy just wanted to be politically correct – a pill should be sweetened. Then, however, it seems that his study belongs to 72% part, doesn’t it?
Trackback responses to this post