Published by
Stanford Medicine

Behavioral Science, Clinical Trials, Research, Science Policy, Stanford News

“U.S. effect” leads to publication of biased research, says Stanford’s John Ioannidis

"U.S. effect" leads to publication of biased research, says Stanford's John Ioannidis

pressuregaugeThe life of a scientist can be filled with pressure – pressure to publish, pressure to obtain funding, pressure to support the people in his or her lab. It’s no surprise that some would-be researchers *cough, me, cough* choose instead to pursue other careers (science writing FTW!).

Now new research by Stanford study-design expert John Ioannidis, MD, DSc, and Daniele Fanelli, PhD, from the University of Edinburgh, suggests that such pressures may lead to more than just sleepless nights. Their results, published today (subscription required) in the Proceedings of the National Academy of Sciences, conclude that, at least in areas of “soft” science that measure behavioral changes, studies with researchers from the United States are more likely to report that the intervention they were testing had an extremely positive outcome than studies whose authors hail from elsewhere. As Ioannidis explained to me last week:

One possible explanation for the pattern that we are seeing is that scientists from the United States are under higher pressure to generate extreme results. This could be for various reasons: to obtain funding, to justify a promotion or to advance one’s career.

Ioannidis isn’t suggesting that scientists are falsifying their results intentionally. Rather, some fields of research are more difficult to quantify than others. “In the behavioral sciences, results are presented with more leeway and creativity than in other ‘hard’ sciences like genetics, when, for example, there’s less room for error when sequencing a gene,” he explained.

For the study, the researchers analyzed more than 1,000 primary outcomes of 82 meta-analyses published in genetics or psychiatry published between 2009 and 2012. And, as Ioannidis points out, things could have changed since that time:

Our study shows what has happened in the past. It’s possible that this phenomenon is becoming more global. Unfortunately, there’s no “investigative pressure meter” we can use to directly compare the policies governing scientific research in various countries.

Previously: Shaky evidence moves animal studies to humans, according to Stanford-led study, Neuroscience studies often underpowered, say researchers at Stanford, Bristol and NIH funding mechanism “totally broken,” says Stanford researcher
Photo by William Warby

Comment


Please read our comments policy before posting

Stanford Medicine Resources: