Testing Alleged Mediumship: Methods and Results
🧐 Skeptical/Critical📌 Appears in:
Plain English Summary
Can professional mediums really talk to the dead? Researchers put five certified mediums to the test with a clever setup: each medium gave hour-long readings for five men, but here's the twist — the mediums couldn't see or hear the sitters (the people getting readings), and nobody knew who any reading was meant for until the scoring was done. This double-blind design shut down any possibility of the mediums picking up subtle cues like body language or appearance. When the results came in, they were pretty devastating for the mediumship camp. Not a single medium scored above chance — none of them could match their readings to the right person. In fact, in 24 out of 25 cases, the wrong person actually rated the reading as more accurate than the person it was intended for! That's a striking finding. The study also uncovered something telling: readings stuffed with vague, general statements (think "someone close to you had health problems") got higher accuracy ratings from everyone, which is a classic sign of the Barnum effect — the same psychological trick that makes horoscopes feel spookily personal. Meanwhile, truly specific statements consistently fell flat. Published in a mainstream psychology journal, this study served as an important reality check against earlier pro-mediumship research that lacked these tight controls, particularly around timing cues that could tip off a savvy medium.
Research Notes
Key skeptical contribution to the mediumship controversy: introduces temporal-cue controls absent from prior studies (including Schwartz et al. 2001) and demonstrates a practical, methodologically rigorous protocol. Published in a mainstream psychology journal, lending weight as a counterpoint to pro-mediumship findings by Beischel and others.
Five professional mediums certified by the Spiritualists Nationalist Union each gave one-hour readings for five male sitters under acoustically isolated, double-blind conditions preventing all sensory leakage. Sitters rated the accuracy of every statement (1-7 scale) from all 25 readings without knowing which was intended for them. Pratt-Birge permutation analyses found no significant results for any individual medium (p = .27 to .89) or combined (p = .63). In 24 of 25 cases, non-target sitters rated the readings higher than the intended sitter. Readings containing more general, diverse statements received higher absolute ratings consistent with the Barnum effect, while highly specific statements consistently received low ratings.
Links
Related Papers
Companion
Also by these authors
Failing the Future: Three Unsuccessful Attempts to Replicate Bem's 'Retroactive Facilitation of Recall' Effect
Of Two Minds: Sceptic-Proponent Collaboration within Parapsychology
Does Psi Exist? Lack of Replication of an Anomalous Process of Information Transfer
More in Mediumship
Intuitive Assessment of Mortality Based on Facial Characteristics: Behavioral, Electrocortical, and Machine Learning Analyses
A Mixed Methods Phenomenological and Exploratory Study of Channeling
People Reporting Experiences of Mediumship Have Higher Dissociation Symptom Scores Than Non-Mediums, But Below Thresholds for Pathological Dissociation
Electrocortical activity associated with subjective communication with the deceased
Neuroimaging during Trance State: A Contribution to the Study of Dissociation
📋 Cite this paper
O'Keeffe, Ciarán, Wiseman, Richard (2005). Testing Alleged Mediumship: Methods and Results. British Journal of Psychology. https://doi.org/10.1348/000712605X36361
@article{okeefe_2005_testing_mediumship,
title = {Testing Alleged Mediumship: Methods and Results},
author = {O'Keeffe, Ciarán and Wiseman, Richard},
year = {2005},
journal = {British Journal of Psychology},
doi = {10.1348/000712605X36361},
}