Last week, Steven Hayward drew attention to a correction in the American Journal of Political Science of a study they had published in 2012. Hayward’s post is entitled Epic Correction of the Decade. I’m thinking it may be the most epic correction of all time. I’ll paraphrase:
“Hey, remember how we said that conservatives are more prone to psychoticism and authoritarianism, while liberals are more prone to neuroticism and wanting to get along with others? It turns out we got that exactly backwards. Sorry, our bad.”
Well, at least they owned up to it. Years later. After a Ph.D. student pointed out the error.
But they’re also arguing (now) that the correlations are so small as to be meaningless, so which way those correlations go doesn’t really matter.
The same study by the same researchers was the basis for three other academic journal articles. According to Retraction Watch, two of those have also published corrections and the last will do so in July.
The trouble with peer review in the social sciences is that all the peers doing the reviewing share the same worldview and prejudices as the “researchers” being reviewed.
Oh, by the way: one of the authors said the surveys for this “research” were conducted in the 1980s and 1990s, and were combined and analyzed in the 2000s. So these “scientists” worked on this stuff for three decades. And your tax dollars paid for it.