Part I describes a multi-laboratory experiment to assess the hypothesis that changes in the behavior of random event generators (REG) might be correlated with the coordinated activity of large numbers of people. A specific prediction was made that the composite of data taken during a five-minute period of meditation shared by people all over the world would show a deviation from its theoretically expected mean value.

Data were recorded from 14 independent REG systems in seven different locations, all in the US or Europe. The composite result shows a meanshift of slightly less than two standard deviations (Z = 1.984) with a corresponding probability of P = 0.047. This result provides evidence for the hypothesized correlation of the REG results with the global meditation. More impressive, however, is the size of the anomalous effect, which is several times greater than that found in typical laboratory experiments using similar equipment.

Part II describes a conceptual replication that was implemented during a different global event some months later. In this case there was no evidence for an anomalous correlation. Differences between the two applications are discussed as factors that may be relevant to the significantly different outcomes.