Probably everyone reading this article has heard of at least one of food science and eating behavior expert Professor Brian Wansink’s studies, and many will have used them, particularly to develop healthier eating habits. Now six of Wansink’s papers have been retracted by the journals in which they were published, adding to seven previous retractions, after the validity of some of his most popular work came into question.
Wansink is a nutrition researcher at Cornell University, with a talent for proposing research topics that catch the popular imagination. You’ve probably heard his claim that if you don’t want to buy high-calorie foods, don’t do your grocery shopping hungry, or that the size of your plate influences how much food you will take when serving yourself. Outside nutrition, Wansink’s work on how using graphs adds credibility to your arguments made it into a larger story IFLScience ran.
After some unwise statements on a blog post, however, Wasnick’s methods came under scrutiny from fellow researchers and journalists.
In February, Buzzfeed claimed Wansink was encouraging his junior researchers to analyze data in a way that made it much more likely they would find interesting results, called p-hacking. When the results didn’t match what Wasnick wanted, the allegation went, he had his team look at subsections of the data until something he could publish came up. Emails Buzzfeed acquired suggested that if an effect couldn’t be found among a population as a whole, Wansink wanted it tested among just men, just women, only alcohol drinkers and so on.
The approach, mocked by XKCD, is one scientists are trained to avoid, but can sometimes prove irresistible, particularly when your next grant depends on having a long list of publications.
In May the Journal of the American Medical Association (JAMA) and journals in the JAMA network published expressions of concern over six of Wansink’s papers while requesting more information. The list includes several of Wasnick’s most famous pieces of work.
JAMA sought assurance from Cornell on the quality of the work, but was told, “We regret that, because we do not have access to the original data we cannot assure you that the results of the studies are valid.” These papers have now been withdrawn.
In Wansink’s case, it is debatable how much this matters. Certainly, every scientist who missed out on the grant funding Wansink received will feel aggrieved, but there may be no serious consequences if people take Wansink’s advice to pre-order school lunches.
On the other hand, life-and-death studies on the effectiveness of medication are sometimes also alleged to be similarly p-hacked. If Wansink’s tale bolsters efforts to tackle this, it could make science stronger.
Update: Buzzfeed reports Wansink has resigned from Cornell following a finding of scientific misconduct. He will continue working there until June 2019, assisting with a review of his research.