A quantitative digest of the ESP-Nexus structured study database for Field-RNG & the Global Consciousness Project: 28 studies spanning 1996–2025, with every figure traceable to its source. ESP-Nexus reports what the studies found and takes no position on whether the effects are genuine.
The Global Consciousness Project (GCP) and related field-RNG research spans work published from 1996 to 2026 across 36 studies. A 2008 meta-analysis by Nelson pooled k=247 events and reported a cumulative effect size of 0.313 (z/sqrt(n)), which proponents treat as evidence of a consistent small-to-moderate deviation in random-number generators during globally shared events. Several individual studies report large z-values, including z=7.31 appearing across multiple independent reports (2021 Williams, 2023 Plonka, 2024 Nelson) and a 2024 Nelson result of p=2.7×10−6, suggesting that at least some analyses detect signals well beyond conventional thresholds.
However, the evidence base is not uniform. Two results report null or below-chance effects, including a 2026 Radin z of -2.97 and a 2025 Radin z of -4.9, indicating negative deviations from chance. This directional disagreement across studies, combined with notable heterogeneity in z-values and the recurring question of whether GCP event selection and analysis protocols are sufficiently pre-registered, means the literature remains contested. No independent consensus has resolved whether the positive results reflect a genuine anomaly or methodological artifacts.
Pooled meta-analytic effects
- Nelson (2008): ES 0.313 (z_over_sqrt_n) (k = 247)
The evidence
| Paper | Reported finding | Effect / significance | Basis |
|---|---|---|---|
| Radin (2025), Journal of Scientific Exploration [source] | Principal Components Analysis composite at midnight. | z = -4.9, p = 4.8 × 10−7 | — |
| Nelson (2025), Journal of Scientific Exploration [source] | FieldREG combined result – all 5 segments, Khufu Subterranean Chamber. | p = .013 | N = 5883 |
| Nelson (2024), Journal of Scientific Exploration [source] | All Formal and Extensions. | p = 2.7 × 10−6 | k = 80 events/tests |
| Nelson (2024), Journal of Anomalous Experience and Cognition [source] | Compounded formal GCP experiment across 500 pre-registered events. | z = 7.31 | k = 500 events/tests |
| Cavanaugh et al. (2022), Studies in Asian Social Science [source] | Main analysis: US homicide rate trend change 2007-2011 vs baseline. | p < 1 × 10−4 | — |
| Nelson (2020), Journal of Scientific Exploration [source] | GCP formal series bottom-line meta-result. | z = 7.0 | k = 500 events/tests |
| Radin et al. (2017), IONS Report (Institute of Noetic Sciences) | Combined across 5 experiments — Man Burn. | z = 3.25, p = .001 | k = 5 |
| Nelson (2014), Journal of International Society of Life Information Science (J. Intl. Soc. Life Info. Sci. / ISLIS) | Composite of all formal GCP hypothesis tests. | z = 7.0, p ~ 1 × 10−12 | k = 456 events/tests |
| Shimizu et al. (2012), Zeitschrift fuer Anomalistik / Journal of Anomalistics | Combined experimental condition – cumulative Chi-Z. | z = 2.08, p = .038 | k = 6; N = 230 |
| Shimizu et al. (2012), Journal of Scientific Exploration | Audience size vs Random Streamer CZ at stadium. | ES 0.31, p = .002 | k = 94 events/tests; N = 94 |
| May et al. (2011), Journal of Scientific Exploration | GCP formal events – total Stouffer's Z. | z = 5.81 | k = 300 events/tests |
| Bancel (2011), Journal of Scientific Exploration | Theil-Sen robust estimator, full dataset – Monte Carlo test of DAT. | z = 1.98, p = .024 | — |
| Shimizu et al. (2010), Journal of Scientific Exploration | RPG102 Z-chi, view condition vs chance. | z = 2.01, p = .04 | — |
| Wilson et al. (2010), Journal of Scientific Exploration | Overall pooled field RNG across 14 sessions. | p = .559 | k = 14 events/tests |
| Nelson (2008) | GCP composite over 247 formal replications. | ES 0.313, z = 5.121 | k = 247 events/tests; N = 247 |
| Bancel et al. (2008), Journal of Scientific Exploration | Formal experiment cumulative network z-score. | z = 4.55, p = 3 × 10−6 | k = 236 events/tests |
| Mason et al. (2007), Journal of Scientific Exploration | Experiment A – Meditation Hall. | z = -4.726, p = 1.1 × 10−6 | N = 11360 |
| Nelson (2006), Proceedings of Presented Papers (The Parapsychological Association Convention 2006) | Devvar – composite signal-averaged across 8 years, permutation combined statistic. | p = .026 | 8 studies |
| Nelson et al. (2006), AIP Conference Proceedings | 86 NA + Eurasia quakes R>=6, focused covar dip permutation analysis. | z = 3.852 | k = 86 events/tests |
| Hagel et al. (2005), Zeitschrift fuer Anomalistik / Journal of Anomalistics | Headline self-reported pattern: CD deviations frequently time-coincident with local emotionally charged events. | — | N = 151 |
What is unsettled
- Reported directions disagree across studies (some positive, some null or below-chance).
- 1 result(s) report a null or below-chance (negative) effect.
Source: ESP-Nexus structured study database, 28 studies. 28 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.