The Myth of Publication Bias in Psi Research: A Comparison Between Parapsychology and Mainstream Psychology

Abstract

This study investigates the persistent claim that publication bias contaminates (indeed, possibly inflates) the scientific evidence for parapsychological phenomena, specifically extrasensory perception (ESP). We analyzed 165 published studies (243 experiments) from recent ESP meta-analyses and 40 preregistered confirmatory experiments from the Koestler Parapsychology Unit Registry. We compared these two datasets to datasets from the field of mainstream psychology. Our primary measure was the percentage of peer-reviewed experiments reporting null statistically significant outcomes (i.e., p > 0.05). Results indicate that the rate of published experiments with null outcomes in ESP research is considerably higher than that observed in psychology, for both non-preregistered and preregistered experiments. While the publication bias is substantially smaller, suggesting more balanced reporting of parapsychological outcomes compared to mainstream psychological research, we cannot conclusively show that the statistical evidence for ESP is not an artifact of that relatively minimal bias, but a number of earlier tests on the file-drawer problem do undermine that assumption.