“Joy of Cooking” Authors Question Brian Wansink Study

Brian Wansink, director of the Food and Brand Lab at Cornell University, is facing criticism from a couple that owns the rights to “The Joy of Cooking,” which Wansink asserted had increased calorie counts of its recipes by an average of 44 percent since its first publication in 1936. The New Yorker reports that John Becker and his wife were “blindsided” when Wansink published “The Joy of Cooking Too Much” in the Annals of Internal Medicine in 2009, but they assumed his findings were correct. At the time, they posted a response on the cookbook’s website saying that out of 4,400 recipes, Wansink had analyzed 18.

Becker later saw a cartoon commissioned by Cornell to appear with Wansink’s original study, and he decided to check Wansink’s results, apparently finding numerous recipes that contradicted Wansink’s findings. Becker forwarded his research to James Heathers, a behavioral scientist at Northeastern University, who reportedly found issues in the 2009 study: Wansink had compared recipes on the basis of serving size despite 10 of the 18 recipes lacking specific serving sizes, and he also compared recipes that created different dishes despite bearing identical titles. For example, Heathers told The New Yorker, one of Wansink’s “egregious calorie gainers” was a recipe for gumbo; the 2006 version is thickened with roux and contains sausage and chicken, while the 1936 version was a clear soup of chicken and vegetables.

Wansink reportedly told the magazine that the published data was an abbreviated version of his full study and that he removed “several key elements” before publication at the request of the journal’s editors. Several of Wansink’s studies have been retracted, and Cornell has reportedly begun a formal investigation into his work.