Mr LaCour’s Unbelievable Findings
On my way to work I was listening to the podcast This American Life. They reported a recent study which found that canvassers could change voters’ minds about gay marriage by sharing personal stories about how it affects them. I found this really exciting because the implications for politics were huge. It was great to see the Social Sciences producing robust and applicable research. In addition, the findings were published in the high impact journal, Science.
However, a week or so later issues with the study came to light. A pair of researchers from U. C. Berkeley took a look at the raw data after they had trouble conducting a follow-up study using similar methods. There were patterns that indicated fabrication – the trends were just too regular. Real people do not behave neatly. The original study’s principal investigator, Professor Green, issued a request for a retraction to the journal. The graduate student who is said to be responsible, Mr LaCour, has yet to give a full response.
On closer inspection, the canvassing itself does appear to have happened. Hundreds of highly trained individuals with personal stories about how gay marriage would affect their lives went door-to-door in California. The fabricated part was the follow-up survey, which purported to find that individuals’ opinion had changed, and stayed changed. In reality, these follow-up surveys never happened. Mr LaCour created the data himself to produce an exciting result which flew in the face of all other previous research.
This is illustrative of a big problem in science. A huge amount of work had gone into the study, and a null finding would almost certainly not gain any interest and would not be published in a prestigious journal. This would mean lots of ‘wasted’ time and money with no return. Researchers are under a lot of pressure to produce novel and exciting findings because of this. When it comes to the importance of publishing in securing a job, Dr. Oransky, co-founder of the website Retraction Watch and quoted in the New York Times, put it perfectly: “They don’t care how well you taught. They don’t care about your peer reviews. They don’t care about your collegiality. They care about how many papers you publish in major journals.”
It’s not all bad news, though. The fact that the fabrication was picked up on is a good example of self-correcting science in action. We can also learn some valuable lessons about recognising good science, even if it is not flashy, and even if nothing is found.
Image: ‘Tell me again’ under creative commons license