A quantitative digest of the ESP-Nexus structured study database for Precognition: 44 studies spanning 1974–2025, with every figure traceable to its source. ESP-Nexus reports what the studies found and takes no position on whether the effects are genuine.

Across four meta-analyses and dozens of primary studies spanning 1988 to 2025, the quantitative record on precognition is genuinely mixed. Honorton (1989) pooled 309 forced-choice experiments and reported a z of 11.41, while Bem (2016) synthesized 90 experiments and found a Hedges g of 0.09. By contrast, Steinkamp (1998) analyzed 22 forced-choice studies and obtained a z-over-root-n of only 0.01, essentially indistinguishable from chance. Among individual studies, results range widely: Bancel (2022, 2025) reported z-scores of 3.7 and 4.0, Mossbridge (2023) found p less than 1×10-6, and Cameron (2024, 2025) reported a Cohen's d of 0.36 in both years, whereas Walleczek (2025) returned p = 0.979, Alibalaei (2025) found d = -0.027, Muhmenthaler (2022) found d = 0.001, and Maier (2020) returned p = .76. Fourteen results in the dataset are null or below chance, and effect-size heterogeneity across studies is substantial, meaning no single summary figure characterizes the literature cleanly.

Pooled meta-analytic effects

The evidence

PaperReported findingEffect / significanceBasis
Bancel et al. (2025), Journal of Anomalous Experience and Cognition [source] [audit]Tryout data – All cohorts pooled.z = 4.0, p = 3.1 × 10−5, hit rate 0.5059N = 1789
Mossbridge et al. (2024), Journal of Anomalous Experience and Cognition [source] [audit]Exp 2 free-response overall target precognition.ES 0.22, p < 2 × 10−4, hit rate 0.35N = 307
Muhmenthaler et al. (2022), Psychology of Consciousness: Theory, Research, and Practice [source] [audit]Experiment 1 – Retroactive Affective Priming.ES 0.001, p = .512N = 36
Walleczek et al. (2025), PLoS One [source]Study 1 – pure-session erotic trials, preregistered confirmatory one-sided.p = .979, hit rate 0.4948N = 37836
Walleczek et al. (2025), Preprint V1 (under peer review, dated 06/27/2025; journal not specified)Study 1 – confirmatory, erotic trials, pure session.p = .979, hit rate 0.4948N = 37836
Alibalaei et al. (2025), Journal of the Society for Psychical ResearchStudy 1 – 40-day yoga course, pre vs post precognition.ES -0.027, p = .783, hit rate 0.251N = 4160
Boyle (2025), Journal of Scientific Exploration [source]ChatGPT-4.1-mini 100-trial precognition task.ES 0.28, p = .005, hit rate 0.32N = 100
Sheldrake et al. (2025), Journal of Anomalous Experience and Cognition [source]Overall precognition condition – standardized effect size.ES 0.18, p = .17k = 3
Radin et al. (2023)Exp 1 – hidden-task slope.z = 2.267, p = .02N = 85792278
Mossbridge (2023), Journal of Anomalous Experience and Cognition [source]EEG presentiment – pre-stimulus prediction of future button-press response.p < 1 × 10−6N = 4000
Bancel (2022), Unpublished proposal/report (IMI institute)Psi@Home preliminary variance test.z = 3.7, p = 1 × 10−4, hit rate 0.5N = 1160
Schlitz et al. (2021), Journal of Scientific Exploration [source]Exp 1 confirmatory: retroactive priming, full sample.ES 0.03, p = .22N = 493
Maier et al. (2020), PLOS ONE [source]Cross-lab random-effects meta-analysis.p = .765 studies; N = 2004
Schmidt (2019), Journal of Parapsychology [source]Experiment 1 – combined total.ES 0.0253, z = 6.36, p < 2 × 10−9k = 3 events/tests; N = 63066
Jolij et al. (2019), Journal of Scientific Exploration [source]Connectivity <|R|> – Amsterdam.p = .018N = 61
Mueller et al. (2019), Zeitschrift fuer Anomalistik / Journal of Anomalistics [source]Overall ARV hit rate over 48 valid DAX predictions.ES 0.56, z = 3.897, p = 2.3 × 10−5, hit rate 0.7916N = 48
Katz et al. (2018), Journal of Scientific ExplorationProject Firefly overall – 177 daily aggregate predictions.hit rate 0.48N = 177
Kugel (2018), Zeitschrift fuer Anomalistik / Journal of Anomalistics [source]Program prognosis module – overall.z = 3.27, p = 6 × 10−4N = 3713
Alvarez (2018), Journal of Scientific Exploration [source]Active block – Experimental trials: distance to to-be-lit vs not-to-be-lit LED.p = .035N = 25
Vernon (2018), Journal of Parapsychology [source]Main study: Precall vs.ES 0.32, p = .021N = 1980
Source: ESP-Nexus structured study database (44 studies; the table shows the 20 highest-ranked). ESP-Nexus reports what each study found and takes no position on whether the effects are genuine.

What is unsettled

Source: ESP-Nexus structured study database, 44 studies. 44 born-digital figures; OCR-derived figures awaiting source-verification are excluded from this digest. Effect sizes are shown only where a standardized metric is recorded.