A quantitative digest of the ESP-Nexus structured study database for Macro-PK & Dice: 12 studies spanning 1955–2023, 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 largest pooled estimate in this body of work comes from Radin's 1991 meta-analysis of 148 studies, which returned an effect size of 0.0122 (expressed as z over square-root-n), small in magnitude but drawn from a sizable trial base. Individual studies span a wide range of outcomes: Forwald produced a z of 5.17, Thompson reported p = .50 × 10−04, McConnell reported approximately 1 in 500 (~.002), Puthoff p = 0.004, and Dunne p = .017, all nominally significant. Wahbeh (2018) returned z = 2.25 and Rock (2015) a hit rate of 0.8.

Against these, Wilson (2010) yielded p = .559, offering no evidence of an effect, and Auerbach (2023) produced a Pearson r of -0.397, a below-chance result. Direction of effects therefore disagrees across studies, and the range from strongly positive to null to negative signals meaningful heterogeneity. No consistent moderator has been identified in this sample to explain that variability, leaving the evidence base genuinely unsettled.

Pooled meta-analytic effects

The evidence

PaperReported findingEffect / significanceBasis
Auerbach et al. (2023), Journal of Parapsychology [source]Temporal decline in daily disturbance counts over time.ES -0.397, p = .001N = 295
Wahbeh et al. (2018), Journal of the Society for Psychical ResearchComposite QNG: channeling vs.z = 2.25, p = .024N = 1133
Gimeno et al. (2017), Journal of Scientific ExplorationScale-based PK discrimination — 23 meetings.N = 23
Rock et al. (2015), Journal of Scientific ExplorationGimeno: Red Lights Sitter-Group Sessions.hit rate 0.8N = 15
Wilson et al. (2010), Journal of Scientific ExplorationOverall pooled field RNG across 14 sessions.p = .559k = 14 events/tests
Vilenskaya et al. (1993), Subtle Energies & Energy Medicine JournalSperansky – operator distant effect on poisoned mice.1 studies; N = 24
Dunne et al. (1992), Journal of Scientific ExplorationRemote Random Mechanical Cascade, right-left split.p = .017k = 26 events/tests; N = 10
Radin et al. (1991), Journal of Scientific ExplorationFull database – experimental studies.ES 0.0122, z = 18.2148 studies; N = 2592817
Thompson (1991), Subtle Energies & Energy Medicine JournalBin mean v4 RT-LT difference – original anomaly.p = 5 × 10−5N = 1131
Puthoff et al. (1975)Superconducting differential gradiometer – activity/no-activity correlation.p = .004k = 13 events/tests; N = 130
McConnell (1955)Decline / quadrant effect — within-page score decline.p = .1N = 170000
Forwald (1955)Two-area placement system – Total.z = 5.17N = 71500
Source: ESP-Nexus structured study database (12 studies). 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, 12 studies. 12 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.