Meta-Analysis of Mind-Matter Experiments: A Statistical Modeling Perspective

Are there relationships between consciousness and the material world? Empirical evidence for such a connection was reported in several meta-analyses of mind-matter experiments designed to address this question. In this paper we consider such meta-analyses from a statistical modeling perspective, emphasizing strategies to validate the models and the associated statistical procedures. In particular, we explicitly model increased data variability and selection mechanisms, which permits us to estimate “selection profiles” and to reassess the experimental effect in view of potential other effects. An application to the data pool considered in the influential meta-analysis by Radin and Nelson (1989) yields indications for the presence of random and selection effects. Adjustment for possible selection is found to render the, without such an adjustment significant, experimental effect non-significant. Somewhat different conclusions apply to a subset of the data deserving separate consideration. The actual origin of the data features that are described as experimental, random, or selection effects within the proposed model cannot be clarified by our approach and remains open.