A meta-analysis of mass-media tests of extrasensory perception

Abstract

Some scientists have argued that recent meta-analyses of many different types of parapsychological study suggest that extrasensory perception (ESP) might exist, albeit as a small effect. Large-scale ESP experiments conducted via newspapers, magazines, radio and television tan generate a huge number of guesses and offer researchers a way of quickly obtaining enough data to discover reliably whether such small effects actually exist. The experimental conditions used in mass-media ESP studies are almost identical to most national lotteries (i.e. large numbers of people sitting at home attempting to guess the identity of a distant target) and so positive results from such studies would challenge the notion that lotteries are unpredictable. Meta-analysis of eight ESP studies conducted via the mass media, representing over 1.5 million individual trials, indicate a very low, negative effect size (z/N-1/2 = -.0046) whose overall cumulative outcome did not differ significantly from chance expectation (Stouffer z = -1.60). The paper discusses the implications of these results for the debate about the existence of ESP and its practical implications for lottery organizers.