Study featured in episode 048 retracted

One of the authors of a study featured prominently in a recent episode of the You Are Not So Smart podcast has asked for a retraction after learning his co-author may have falsified most of the survey data included in the research.

Donald Green, a professor of political science at Columbia University, decided to retract the study after his co-author, UCLA graduate student Michael LaCour, failed to produce the computer files containing the survey results upon Green’s request. LaCour claimed those files had been deleted accidentally, but it was later revealed they had never been recorded, said Green. Green’s confrontations with LaCour over the matter were accompanied by a witness.

The original research was published in the journal Science under the title “When contact changes minds: An experiment on transmission of support for gay equality.” In the study, canvassers spoke to opponents of same-sex marriage in an attempt to change those voters’ attitudes and opinions. In recorded interactions, those opponents reported to canvassers that they had flipped on the issue in the course of the conversations. Those attitudes were then supposedly tracked and re-evaluated over time using online surveys, but it now appears those surveys were never conducted.

The matter was brought to Green’s attention a few days earlier by Berkeley researchers attempting to replicate LaCour’s work. Those researchers were unable to reproduce the high rate of response he reported – the number of people who agreed to fill out the surveys out of the total number of people approached. That discrepancy led to an investigation that uncovered previous survey results which seemed similar to those presented by LaCour. The results presented were most likely modified versions of previous survey results copied from a different study, said the researchers. Those findings can be viewed at this link.

Green expressed sadness and embarrassment over the matter on Wednesday and said that he feels the people most hurt in this are the canvassers who agreed to be researched. He added that he believed this retraction should not reflect poorly on the subject matter, and new research should be conducted.

The podcast featuring interviews with LaCour and Green will be re-uploaded later today with that segment removed and replaced with the statements above.