Skip To Content
Incoming Files
← The Anomaly Log
Case File · CJ-2025-0048

Pre-registered precognition test with meditators finds no direct evidence of psi

Date
May 23, 2025
Place
Institut Métapsychique International; online sessions via the Psi@Home platform
Type
Odd Happenings
Status
Concluded
Record
Documented
Filed Under Fringe Science

Can people guess a target before it is chosen? Researchers at the Institut Métapsychique International, a parapsychology research institute, tested the idea again and published the results on May 23, 2025, in the Journal of Anomalous Experience and Cognition.

Peter Bancel, Mario Varvoglis, Jocelyne Boban and Anaïs Bensahra built Psi@Home, an internet platform that lets participants run forced-choice precognition trials at home using custom software. They recruited two groups: experienced meditators, and people selected only for their interest in taking part. Each group completed 80 pre-registered sessions of 20 trials, and the team also collected 90 earlier 'tryout' sessions.

Pre-registration matters here. It means the hypotheses were fixed before the data came in, which guards against fishing for a lucky result. According to the paper's abstract, the main confirmatory test, whether hit rates varied more from session to session than chance would predict, was not confirmed. Only one hypothesis held: the meditators showed more session-to-session variation than the other group (p = .03). The authors state plainly that the pre-registered studies 'found no direct evidence of a psi effect.'

The tryout sessions, using the same people and protocol, did show a strong deviation (p = .00003). The authors suggest participants' attitudes may have differed between the two periods. Skeptics would point to the more ordinary reading: exploratory data that look impressive often shrink once a study is locked down in advance, which is exactly why pre-registration is used.

The abstract does not give raw hit rates. Readers can find the full paper, which is open access, at the link below.

What's disputed or missing

  • Why did the tryout data deviate when the planned sessions did not? The authors' attitude explanation is untested.
  • Would a larger replication find anything? This was a test of a new platform as much as of psi.

Sources

  1. Main
    A Forced-choice Precognition Experiment with Selected Cohorts (Journal of Anomalous Experience and Cognition) The peer-reviewed paper: design, cohorts, p-values and the authors' conclusion.
Back To The Log

Declassified · Direct To You

The CJ Dispatch

New Case Files, Episodes, and Books; Sent When Something Surfaces. Never More Than We Have To Say.

No spam. Unsubscribe anytime.