Discovery Lab Answers

Precognition data

I’ll walk you through what the ESP-Nexus library shows on precognition studies and point you toward the deeper resources.

Overall pattern from forced-choice literature

The sources retrieved for this question include two foundational meta-analyses of forced-choice precognition experiments spanning decades of research.

Honorton and Ferrari (1989) analyzed 309 forced-choice precognition studies published between 1935 and 1987, drawing from nearly two million individual trials across more than 50,000 subjects ^[1]. The meta-analysis found a small but statistically significant overall effect (z = 11.41, p = 6.3 × 10⁻³¹). Thirty percent of the studies achieved significance at the 5% level, contributed by 40 different investigators. The mean effect size across studies was 0.65, with a 95% confidence lower bound of 0.40 ^[1].

A later meta-analysis by Steinkamp, Milton, and Morris (1998) compared forced-choice precognition and clairvoyance outcomes from studies published between 1935 and 1997. A post-hoc analysis found a significant interaction between ESP type and experimental design: precognition trials performed better when interspersed within studies together with clairvoyance trials, whereas clairvoyance trials performed better when the two types were run as separate experiments ^[2].

Current research directions

Beyond the historical meta-analyses, the library profiles researchers developing new precognition methodologies. Edwin C. May has focused on identifying physiological markers of anomalous anticipation and developing theoretical frameworks including a multiphasic model suggesting precognition involves multiple temporal phases rather than a single mechanism. Peter Bancel and Mario P. Varvoglis have both developed Internet-based forced-choice precognition platforms (including the Psi @Home and Sharefield/Selfield protocols) designed to improve replication and participant motivation—though recent preregistered work has reported mixed results, with formal cohorts showing no significant hit-rate increase but tryout sessions showing variance anomalies [see the pages on Bancel and Varvoglis for details].

For a comprehensive synthesis of precognition research, methods, and the gap between data and competing explanations, see the Precognition phenomena page and May’s page on precognition and presentiment mechanisms.