Abstract

This meta-analysis examined 25 ESP studies from 12 chief investigators. No potential moderator variable correlated significantly with ESP effect size (n) under either hypnosis or the comparison condition. There was cumulative ESP-test significance for hypnosis, but this significance may be inflated by nonindependence of study outcomes within chief investigator. For both experimental conditions, chief investigator was associ­ated with significant heterogeneity of outcomes. In ESP analyses based on chief investiga­tor, neither the contrast, nor either of the experimental conditions, showed significance. The sum of scored flaws per study did not correlate significantly with outcomes, but the mean number of such flaws was substantial. For the same-subjects studies suitable for analysis, testing order interacted significantly with the experimental manipulation. The hypnosis-comparison contrast was significant only when the comparison condition pre­ceded hypnosis. This significance was due, substantially, to psi-missing in the comparison condition. These and other findings make it difficult to draw substantive conclusions from the current database.