Discovery Lab Answers

Psychokinesis: what the evidence looks like

Psychokinesis (PK) is the claimed direct influence of mental intention on a physical system without any known physical means. Empirically it splits into two very different research traditions:

• Micro-PK — statistical influence on random systems (random number generators, dice), studied with large trial counts and meta-analysis.
• Macro-PK — large-scale physical effects (object movement, physical mediumship, recurrent spontaneous psychokinesis/RSPK), studied through case investigation rather than controlled trials.

Coverage: this reflects the studies currently in the ESP-Nexus library, not the complete published literature on this question. We have not yet measured how much of that literature the library holds, so treat what follows as a summary of what we hold rather than a settled account of the field.

The quantitative picture (macro-PK dice and related)

The structured evidence retrieved for this question spans experiments from the 1950s to 2023, and it reports a mix of non-comparable metrics — standardized effect sizes, raw hit rates, z-scores, and bare p-values. These cannot be averaged into one “overall PK effect size,” so the table below keeps each metric distinct.

What this shows and does not show. Most retrieved rows report positive, significant departures from chance, some with very large z-values driven by enormous trial counts (Radin 1991 reports z = 18.2 across ~2.6 million trials, but a per-trial effect size of only ~0.012). But the directions disagree: Wilson (2010) pooled 14 field-RNG sessions and reported a null result (p = .559), and Auerbach (2023) reports a temporal decline rather than a straightforward PK hit — the daily counts fell off over 295 events. Note also the very small participant/session counts behind several positive rows (Rock 2015: hit rates of 0.8 and 1.0 on 15 and 10 trials with 5 and 1 participants respectively) — high hit rates on tiny samples carry wide uncertainty.

The meta-analytic dispute over micro-PK on RNGs

The most direct engagement with whether the RNG evidence adds up to a real effect is in the retrieved sources themselves.

What critics argue. Bösch, Steinkamp, and Boller (2006), reviewing the interaction of human intention with random number generators, found that a statistical effect is present in the database but concluded it is best explained by publication bias and the small-study effect — small studies show larger effects, producing funnel-plot asymmetry consistent with selective reporting rather than a genuine PK signal. Pallikari (2023) similarly attributes the scatter of scores to the “small-study effect, the Conformity, and the Publication biases,” and offers a Markov-correlation model in which the broadening of micro-PK scores arises from statistical structure, arguing that a non-paranormal reading “accounts for all evidence” the analysis examined.

What the experimental data show. In a commentary on the Bösch et al. meta-analysis, Radin, Nelson, Dobyns, and Houtkooper agreed that the existing studies provide statistical evidence for psychokinesis and that the evidence is generally of high quality, while contesting the publication-bias interpretation of that same database. The broader review by Williams (2021) reports that proof-oriented empirical evidence for micro-PK on RNG output has accumulated over roughly eight decades, while stating plainly that “there is still very little that is known at present about the processes” that might underlie it. Schmidt (1990) reported that his RNG effects were confirmed by a large number of different experimenters, with effect sizes comparable across his precognition- and psychokinesis-oriented test modes.

Analysis. The disagreement is not primarily about whether the database departs from chance — Bösch et al. and their critics both accept a statistical anomaly is present. It is about interpretation: Bösch, Steinkamp, and Boller (2006) read the funnel-plot asymmetry as publication/small-study bias; Pallikari (2023) offers a Markov-bias model attributing the pattern to unintentional data-shaping; Radin and colleagues read the same evidence as a genuine effect. Williams (2021) frames the open problem as mechanistic — even granting the proof-oriented results, no process account is established.

Theoretical objections

At the level of “could PK be real at all,” George R. Williams (2023) engages the physicist Sean Carroll’s argument that PK (and psi generally) should be dismissed because it conflicts with current understanding of the laws of nature, and proposes an alternative dispositional/quantum-ground framework for causality intended to be consistent with the mind-matter data. This is a conceptual argument, not experimental evidence.

Macro-PK: a separate, non-replicable literature

The macro-PK material on ESP-Nexus is case-based rather than statistical. Rosemarie Pilkington documents and critically analyzes Recurrent Spontaneous Psychokinesis (RSPK) — her preferred term over “poltergeist” — arguing that these dramatic, multiply-witnessed physical phenomena are under-studied relative to laboratory ESP, while noting the central vulnerability that RSPK is state-dependent and non-replicable on demand. Stephen E. Braude approaches physical mediumship and macro-PK as a philosopher and case analyst, weighing evidential value against the recurring problem of fraud and the technical limits of video/infrared documentation. Their pages are the deeper dives:

• Rosemarie Pilkington — Recurrent Spontaneous Psychokinesis and Poltergeist Phenomena
• Stephen E. Braude — Physical Mediumship and Macro-Psychokinesis

For a single audited RNG-PK study, see Maier et al. (2022) — Quantum measurement micro-PK, and the Studies section for more single-study audits.