Skip to main content

Examining Psychokinesis: The Interaction of Human Intention With Random Number Generators—A Meta-Analysis

Contested
Bösch, Holger, Steinkamp, Fiona, Boller, Emil 2006 Modern Era psychokinesis

📌 Appears in:

Plain English Summary

Can you move things with your mind? This landmark meta-analysis, published in one of psychology's most prestigious journals, tackled that question head-on by crunching 380 studies spanning 45 years of people trying to mentally influence random number generators (basically digital coin flips). The verdict? There is a statistically significant effect — but it's breathtakingly tiny, shifting the odds to 50.0286% instead of a perfect 50/50. That's like winning one extra coin flip out of every 3,500. And here's the kicker: smaller studies found bigger effects, which is a classic red flag for publication bias (the tendency for boring null results to stay in file drawers). The researchers ran a clever computer simulation showing that if roughly 1,500 unpublished negative studies existed, it would perfectly reproduce all the patterns in the data. Their blunt conclusion borrowed a phrase from 1962: "not proven." This analysis became the single strongest skeptical challenge to the entire mind-over-matter research program and sparked a heated rebuttal from the field's leading proponents.

Research Notes

The most rigorous independent meta-analysis of the RNG-PK database, published in the APA’s Psychological Bulletin. Its Monte Carlo publication bias simulation provides the strongest skeptical argument against the entire RNG-PK evidence base and directly prompted Radin, Nelson, Dobyns & Houtkooper’s rebuttal. Essential reading for Controversies 4 (PK) and 10 (meta-debate).

A meta-analysis of 380 studies (117 reports, 1959–2004) examining whether human intention can influence true random number generator output. Using both fixed-effects and random-effects models, the analysis found a statistically significant but extremely small overall effect (REM: π = .500286, z = 4.08, p < .001, excluding three outlier studies). However, effect sizes were inversely related to sample size (small-study effect) and extremely heterogeneous (Q = 1508.56, p ≈ 0). A Monte Carlo simulation showed that a simple publication bias model could reproduce all three main findings, requiring approximately 1,500 unpublished null studies. The authors conclude with Girden’s 1962 verdict: “not proven.”

Links

Related Papers

Also by these authors

More in Psychokinesis

📋 Cite this paper
APA
Bösch, Holger, Steinkamp, Fiona, Boller, Emil (2006). Examining Psychokinesis: The Interaction of Human Intention With Random Number Generators—A Meta-Analysis. Psychological Bulletin. https://doi.org/10.1037/0033-2909.132.4.497
BibTeX
@article{bosch_2006_examining,
  title = {Examining Psychokinesis: The Interaction of Human Intention With Random Number Generators—A Meta-Analysis},
  author = {Bösch, Holger and Steinkamp, Fiona and Boller, Emil},
  year = {2006},
  journal = {Psychological Bulletin},
  doi = {10.1037/0033-2909.132.4.497},
}