Brian Wansink and the children’s television character Elmo don’t seem to get along very well. Wansink, a food researcher at Cornell University, has borne the brunt of intense criticism lately from scientists who say his studies of nutrition in children are shaky at best. In September, he and his colleagues retracted a 2012 paper in JAMA Pediatrics. It claimed that putting Elmo stickers on apples made the fruit more appealing to kids choosing school lunch options. But Wansink’s group pulled that article and immediately substituted what they said was a pristine version — a retract-and-replace approach that a small group of journals have employed as a way of cleaning up the scientific record. But now JAMA has retracted the second version, too, after Wansink acknowledged that the article greatly misstated the ages of the children in the study. The move marks the fourth retraction for Wansink, whose institution found his work to be marred by mistakes but not misconduct. Tickle me, data.