In one of the biggest cases of scientific fraud on record, a prominent psychologist has admitted fabricating data in dozens of studies.
Diederik Stapel, who was suspended from his post at Tilburg University in the Netherlands in September, was exceptionally productive. He was responsible for a succession of eye-catching studies on topics including stereotyping and discrimination, the effectiveness of advertising, and the circumstances in which people may perversely prefer negative feedback to praise.
Stapel was suspended after three junior researchers alleged scientific misconduct. But the extent of the problems became known only on Monday, when the concluding that dozens of papers, as well as 14 out of the 21 PhD theses Stapel had supervised, contain fabricated data.
Advertisement
鈥淭his is absolutely horrifying,鈥� says , a social psychologist at the University of Missouri in Columbia. 鈥淲e are talking about research that has major impact in the field of social cognition.鈥� Social cognition is the field of psychology that investigates how our mental processes affect the way we relate to one another.
Hall of shame
In terms of the sheer volume of research implicated, Stapel鈥檚 is one of the worst cases of scientific misconduct on record. The chair of the committee that has examined Stapel鈥檚 work at Tilburg University told Nature that have so far been found to contain fabricated data. If these are all withdrawn, they will exceed the toll of retractions of papers by Jan Hendrik Sch枚n, whose groundbreaking work at Bell Labs, New Jersey, on electronic devices made from organic molecules was found in 2002 to contain .
The ongoing investigation into Stapel鈥檚 work also involves the universities of Amsterdam, where he gained his PhD, and Groningen, where he worked before . Given that he has more than 150 publications to his name, Stapel鈥檚 case may yet come to rival that of cancer researchers Friedhelm Herrmann and Marion Brach of the Max Delbr眉ck Center for Molecular Medicine in Berlin, Germany, who were eventually found to have published 94 papers and book chapters that .
Improbable stats
Tilburg University鈥檚 interim report explains that Stapel often gave his students data to analyse, or approached colleagues with results that addressed questions in which they were interested. The evidence of fabrication comes from anomalies in this material, including suspiciously large experimental effects and a lack of 鈥渙utliers鈥� in the data. Statisticians are now examining further studies for evidence of similar problems.
The case leaves red-faced collaborators cursing themselves for being so trusting. 鈥淚 was duped,鈥� admits of the University of Connecticut in Storrs, who expects to have to retract two papers he published with Stapel examining how 鈥減riming鈥� people by showing them a picture of Albert Einstein can make them feel less intelligent.
The scandal could prove especially damaging for social psychology, likely to titillate the media.
鈥淥ur field is one where a great deal of currency is placed on surprising you,鈥� says Blanton, who is concerned about a dynamic that encourages researchers to progress from 鈥渃ounter-intuitive to cute, to provocative, to 鈥榙efies gravity'鈥�.
Some of Stapel鈥檚 recent work was certainly provocative. A paper published in April in the journal claimed that disordered environments such as littered streets make people more prone to stereotyping and discrimination. Although the Tilburg inquiry has not yet identified the studies that contain fabricated data, Science has already published an about this paper.
In a , Stapel admitted fabricating data and apologised for the damage done to his colleagues and the field of social psychology. 鈥淚 have failed as a scientist,鈥� he said.
