In another bummer of the week, researchers have announced this week that a highly-touted study about the effectiveness of pro-LGBT canvassing was actually comprised of fake data. 

The study, which was published last December, claimed that “a 20-minute, one-on-one conversation with a gay political canvasser could steer voters in favor of same-sex marriage.” What was so striking about the finding was that voters apparently held onto their pro-marriage-equality beliefs even after the conversation, something that doesn’t usually happen when people change their beliefs quickly and because of another person. 

But Donald Green, the study’s senior author, retracted it after learning that co-author Michael LaCour had faked the results. Apparently, canvassers did talk to voters about their opinions on marriage equality, but nobody actually followed up with the voters about how their thoughts had changed. 

The problems came to light after three other researchers tried, and failed, to replicate the study. David Broockman, of Stanford, Joshua Kalla, of the University of California, Berkeley, and Peter Aronow of Yale found eight statistical irregularities in the data set. No one of these would by itself be proof of wrongdoing, they wrote, but all of them collectively suggest that “the data were not collected as described.”

Broockman, Kalla, and Aronow told Green about the paper’s “irregularities” on Saturday, and sent him a summary of their concerns Sunday. According to his retraction letter, Green then contacted Lynn Vavreck, LaCour’s adviser at UCLA, who confronted him on Monday morning. LaCour couldn’t come up with the raw data of his survey results. He claimed that he accidentally deleted the file, but a representative from Qualtrics — the online survey software program he used — told UCLA that there was no evidence of such a deletion. What’s more, according to what Green told Politico, the company didn’t know anything about the project and “denied having the capabilities” to do the survey.

Yesterday, Vavreck asked LaCour for the contact information of the survey respondents. He didn’t have it, and apparently confessed that he hadn’t used any of the study’s grant money to conduct any of the surveys.

This is just awful news, for so many reasons. This study was a huge source of support for the long-known “contact hypothesis,” which posits that "the best way to reduce prejudice against individuals in a minority group is to boost interactions between them and the majority.” Of course, this isn’t the only study to have shown this effect, but it was a big one for our movement.

Not to mention that when one study that appears to show an increase in LGBT support gets retracted, all the vaguely pro-LGBT studies that follow it will be questioned and scrutinized even more heavily by opponents who are already looking for ways to destroy our credibility. This is such a disappointment. 

Here’s a datasets 101 post written by our chief methodologist for all of you aspiring number-crunchers.

Pew Research datasets are widely used by scholars and students. Among the hundreds of survey datasets available, here are some of the most popular and frequently downloaded:

If the 4-per-country rule started now.

Brazil, Romania, Poland, Greece, PRK, and Korea do not have enough scores or all-arounders at this point in time to be given a score.

  • third column is an approximate score for a current four person team.
  • fourth column is an approximate score for a current six person team.
  • fifth column is the difference between the two.

Biggest “winners” (those who would benefit most from the new rule)

  • Switzerland
  • Mexico
  • Italy
  • Canada
  • Sweden

Biggest “losers” (those who would benefit least from the new rule)

  • China
  • Russia
  • Spain
  • Austria
  • Hungary
  • Great Britain

Note: not every gymnast who could make a worlds team has competed yet.