Are Most Published Criminological Research Findings Wrong? Taking Stock of Criminological Research Using a Bayesian Simulation Approach.

Abstract

This study uses Bayesian simulations to estimate the probability that published criminological research findings are wrong. Toward this end, we employ two equations originally popularized in John P.A. Ioannidis' (in)famous article, "Why Most Published Research Findings are False." Values for relevant parameters were determined using recent estimates for the field's average level of statistical power, level of research bias, level of factionalization, and quality of theory. According to our simulations, there is a very high probability that most published criminological research findings are false-positives, and therefore wrong. Further, we demonstrate that the primary factor contributing to this problem is the poor quality of theory. Stated differently, even when the overall level of research bias is extremely low and overall statistical power is extremely high, we find that poor theory still results in a high rate of false positives. We conclude with suggestions for improving the validity of criminological research claims.