The news has been abuzz about the study, or non-study, published and later retracted by a respected peer-reviewed journal, Science.
The Case of the Amazing Gay-Marriage Data: How a Graduate Student Reluctantly Uncovered a Huge Scientific Fraud
The Strangest Thing About Michael LaCour’s Response to the Academic Fraud Allegations Leveled Against Him
LaCour is being investigated. I’ll pay him the courtesy of waiting for judgement, but if these allegations are substantiated – and the evidence so far is damning – the study never happened. It was a fabrication from top to bottom.
It is deeply concerning that this passed through the peer review process, and that Brookman, who brought these issues to light, faced numerous institutional pressures to drop his investigation. (More in the first link above).
If one wishes to improve the quality of science, one must actively try to extinguish all sorts of error, including malfeasance. It appears that this is not encouraged, and is indeed discouraged, by the institutions which we charge with truth-seeking and propagation of truth.
When scientists create their own truths, that’s way beyond the pale. We cannot support this, no matter how noble or well-intentioned the aim.
I support marriage equality advocacy. I sort of want to believe the claims of the study, that canvassing face to face is effective, under these conditions. But if the study cannot be replicated – if, indeed, the actual study never even existed – it is really important to know that.
There have been other cases of such malfeasance. As before, I recommend that we Open-Source all the science, all the data, all the logbooks, the articles, everything needed to audit, verify, replicate and, when needed, to repudiate all scientific studies.
“All problems are shallow when examined by enough eyeballs”, to borrow from Eric Steven Raymond’s Cathedral and the Bazaar essay.