Independent re-analysis of alleged mind-matter interaction in double-slit experimental data
⚡ Contested ↗📌 Appears in:
Plain English Summary
A French researcher got hold of the full 80GB dataset behind a famous claim that human minds can influence light in a double-slit experiment (that's the classic physics setup where photons behave like waves). The original study reported staggeringly strong evidence -- a one-in-a-hundred-million chance of being a fluke. But Tremblay found a critical statistical mistake: the researchers trimmed their data before running their significance tests instead of after, which made results look roughly 100,000 times more impressive than they actually were. When he re-ran everything properly, correcting for the fact that the team had checked dozens of different time points and light patterns, the headline-grabbing result evaporated. There were still small shifts in the predicted direction, but nothing that cleared the bar for statistical significance. The full data and code are publicly available, making this a model of transparent skeptical replication.
Research Notes
Landmark skeptical paper that pinpointed specific statistical error in prominent psi experiment. Tremblay (CNRS researcher) obtained full 80GB dataset and Matlab codes from Radin, conducted thorough independent re-analysis. Shows how methodological choices (trimming procedure, data combination, time lag selection) can inflate significance by orders of magnitude. Published in PLoS ONE with open data on OSF. Directly responds to Radin et al. (2016 Physics Essays); pairs with Walleczek & von Stillfried (2019 Frontiers in Psychology) which independently identified same statistical flaw using meta-experimental protocol. Essential for understanding the double-slit controversy.
Independent re-analysis of Radin et al.'s 2-year double-slit mind-matter interaction dataset (8,655 sessions). Identifies erroneous trimming procedure in original analysis (trimming before bootstrapping instead of after) that produces uncontrolled false positives and underestimates p-values by ~5 orders of magnitude. Re-analysis with proper statistical methods finds shifts in fringe visibility in the predicted direction but not statistically significant (p>0.05 after Holm-Bonferroni correction for multiple comparisons across 26 time lags and 19 fringes). Robustness confirmed across trimming intensities, session length thresholds, and four fringe visibility estimation methods. Original claim of 5.72σ evidence (p=1.05×10⁻⁸) not supported; actual evidence much weaker (p~10⁻³ under original analytical choices).
Links
Related Papers
Companion
- False-Positive Effect in the Radin Double-Slit Experiment on Observer Consciousness as Determined with the Advanced Meta-Experimental Protocol — Walleczek, Jan (2019)
- Response: Commentary: False-Positive Effect in the Radin Double-Slit Experiment on Observer Consciousness as Determined With the Advanced Meta-Experimental Protocol — Walleczek, Jan (2020)
- False-Positive Effect in the Radin Double-Slit Experiment on Observer Consciousness as Determined With the Advanced Meta-Experimental Protocol — Walleczek, Jan (2019)
More in Psychokinesis
Observer Influence on Quantum Interference: Testing the von Neumann-Wigner Consciousness-Collapse Theory
New Year's Eve as a Case Study in Experimental Metaphysics: Exploring Global Consciousness in Random Physical Systems
Anomalous Entropic Effects in Physical Systems Associated with Collective Consciousness
Psychophysical Interactions with Electrical Plasma: Three Exploratory Experiments
Psychophysical Interactions with a Double-Slit Interference Pattern: Exploratory Evidence of a Causal Influence
📋 Cite this paper
Tremblay, Nicolas (2019). Independent re-analysis of alleged mind-matter interaction in double-slit experimental data. PLoS ONE. https://doi.org/10.1371/journal.pone.0211511
@article{tremblay_2019_reanalysis_double_slit,
title = {Independent re-analysis of alleged mind-matter interaction in double-slit experimental data},
author = {Tremblay, Nicolas},
year = {2019},
journal = {PLoS ONE},
doi = {10.1371/journal.pone.0211511},
}