Confessions of wrongness in academic research should be unsurprising. (To be clear, being wrong in a prediction is different from making an error. Error, even if committed unknowingly, suggests sloppiness. That carries a more serious stigma than making a prediction that fails to come true.) Anyone who has a passing familiarity with the social sciences is aware that, by and large, we do not get an awful lot of things right. Unlike that of most physical and natural scientists, the ability of social scientists to conduct experiments or rely on high-quality data is often limited. In my field, international relations, even the most robust econometric analyses often explain a pathetically small amount of the data’s statistical variance. Indeed, from my first exposure to the philosopher of mathematics Imre Lakatos, I was taught that the goal of social science is falsification. By proving an existing theory wrong, we refine our understanding of what our models can and cannot explain.
And yet, the falsification enterprise is generally devoted to proving why other scholars are wrong. It’s rare for academics to publicly disavow their own theories and hypotheses. Indeed, a common lament in the social sciences is that negative findings—i.e., empirical tests that fail to support an author’s initial hypothesis—are never published.