A 2005 study suggesting a link between dancing and sexual selection in humans was probably based on faked data, according to a recently released review by Rutgers University. The study, which asked Jamaican teenagers to rate the dancing skills of their peers, joins a growing list of debunked media-friendly studies. But here’s something unusual: The investigation into the study’s veracity, which claimed that more physically symmetrical teenagers were rated higher by their peers, was apparently initiated by the paper’s lead author.
Nature used the study on the cover of its magazine back in 2005, under the catchy headline “Fascinating Rhythm.” And while the publication still hasn’t retracted the study itself (that’s still under review), one of the magazine’s reporters has chronicled evolutionary biologist Robert Trivers’ quest to get his own paper retracted. We’ll let them explain:
Trivers began to suspect that the study data had been faked by one of his co-authors, William Brown, a postdoctoral researcher at the time. In seeking a retraction, Trivers self-published The Anatomy of a Fraud, a small book detailing what he saw as evidence of data fabrication. Later, Trivers had a verbal altercation over the matter with a close colleague and was temporarily banned from campus. An investigation of the case, completed by Rutgers and released publicly last month, now seems to validate Trivers’ allegations. Brown disputes the university’s finding, but it could help to clear the controversy that has clouded Trivers’ reputation as the author of several pioneering papers in the 1970s.
Trivers’ body of work includes a long-term study of asymmetry in Jamaican children and teenagers, Nature explains. That wider set of research doesn’t seem to be in question here—the dancing study was just one part of that work. Brown, according to Rutgers, altered the asymmetry measurements of the study’s participants to force a larger correlation between symmetry and dancing skill. For what it’s worth, Trivers and others have claimed that the study’s unaltered data set does appear to show a weak correlation between symmetry and peer ratings, but nothing nearly as dramatic as what’s in the published study.
Meanwhile, Brown has defended himself by arguing that Rutgers made a mistake in comparing his data set to the one held by Trivers’ group, as opposed to another “original” set of data from the study. In any case Nature notes that the investigation has been sent to the National Science Foundation, which gave Trivers and Lee Cronk, his research partner at the time, a $25,000 grant for the motion-capture technology used in the study.
Read the full story at Nature.