A study published in the journal Science said that respondents were likelier to form lasting and favorable opinions regarding gay marriage if they were surveyed by gay canvassers. However, the study is being retracted by one of the authors because of doubts over its data.
At the time they published their results, Michael LaCour from the University of California, Los Angeles and Donald Green from Columbia University were initially skeptical so they recreated their experiment. Results were the same so they concluded that the study was indeed valid. Their work received nationwide coverage, and many other researchers were impressed, prompting extensions of the study to be carried out.
Joshua Kalla and David Broockman from the University of California, Berkeley built upon the results of LaCour and Green's study, even corresponding with the researchers for advice on design details. When Kalla and Broockman actually began a pilot of the extension, however, it was revealed that their resulting response rate was lower than what LaCour and Green reported.
The UC Berkeley researchers also found other irregularities in the original study, which they sent to Green for review. The author agreed that their research should be retracted unless LaCour was able to provide evidence disproving Kalla and Broockman's work.
"Michael LaCour's failure to produce the raw data coupled with the other concerns noted above undermines the credibility of the findings," wrote Green in a retraction letter to Science. He added that he was deeply embarrassed by how events turned out and would like to apologize to readers, reviewers and editors of the journal.
Science also released an "Editorial expression of concern" to inform readers about the serious issues raised regarding the validity of LaCour and Green's study.
"Science is urgently working toward the appropriate resolution, while ensuring that a fair process is followed," said Marcia McNutt, editor-in-chief of the journal.
She added that no peer review is completely perfect, but it is a good thing that science is capable of self-correction. As studies are scrutinized, confirmed or rebutted, science advances.
David Fleischer, project director for the Los Angeles LGBT Center Leadership LAB, said their group was shocked and disheartened by the incident. However, they support Green's decision to retract the study and are grateful that problems within the research have been revealed.
For the original study, around 9,500 voters from the Los Angeles County were surveyed.
Photo: Jose Antonio Navas | Flickr