Testing for Telepathy in Connection with E-mails

The primary email-channel extension of Sheldrake and Smart's telephone telepathy program. The videotaped series with covered computer screens and blind independent evaluation of tapes addresses the main cheating hypotheses for unfilmed trials. Hit rates did not decline from unfilmed (43%) to filmed (47%) conditions, arguing against systematic fraud. Introduces e-mail as a more easily standardized and automatable paradigm for future psi research; directly motivated the automated follow-up (sheldrake_2009_automated_email).

This paper reports two series of experiments testing whether participants could correctly identify which of four potential e-mailers was about to send them an e-mail before receiving it. The design was modeled on prior telephone telepathy studies: one minute before each designated trial time, participants e-mailed the experimenter with their guess; the randomly selected e-mailer then sent their message at the exact designated time, copying the experimenter, providing time-stamped records establishing that guesses preceded messages. In Series 1 (unfilmed), 50 participants completed 552 trials; there were 235 hits (43%), far exceeding the 25% chance baseline (z = 9.49, p = 2 x 10^-19, Cohen d = 0.42; 95% CI 38%-47%). Forty-three of 50 participants scored above chance (expected 24 by chance; z = 5.54, p = 2 x 10^-8). An additional 48 partial-series participants achieved 34% hits (z = 2.94, p = 0.002). In Series 2 (videotaped), five high-scoring participants from Series 1 each completed 30 trials under continuous video surveillance with computer screens covered between trials; in 137 filmed trials there were 64 hits (47%; z = 5.77, p = 3 x 10^-8, d = 0.50). Four of five participants scored individually above chance. Familiar e-mailers produced higher raw hit rates than unfamiliar ones; among participants with two unfamiliar e-mailers and after correcting for response bias, the familiar-unfamiliar difference was statistically significant (z = 3.37, p = 0.0004). Some participants achieved high hit rates with senders thousands of miles away, including one participant whose senders were in Hong Kong (~6,000 miles). Results replicate Sheldrake and Smart's telephone telepathy findings and favor a telepathy interpretation over pure clairvoyance or precognition hypotheses, given the significant familiarity advantage.