Discovery Lab Answers

Remote Viewing Data — What the ESP-Nexus Library Shows

Coverage note (required): The evidence table below reflects studies currently held in the ESP-Nexus library. The library’s share of the full published literature on remote viewing has not been measured, so everything that follows is a summary of what the library holds — not a settled account of the field as a whole.

The 21 studies in the evidence base span roughly five decades (1975–2026), use several incompatible metrics, and produce a genuinely mixed picture. Because the metrics cannot be averaged or pooled across rows, the summary below describes patterns within coherent groups rather than offering a single bottom-line figure.

Directions of Effect Across the Evidence Base

The majority of rows report a positive direction — performance above chance — but three report null or below-chance results, and between-study heterogeneity is high in at least one pooled analysis. That spread is real and must be read as part of the picture, not as noise to be set aside.

Positive results appear across multiple designs and laboratories:

• The largest systematic evidence comes from Tressoldi (2023) — the Katz and Tressoldi meta-analysis of 36 studies (40 effect sizes, 1974–2022) — which found a positive standardized effect using a random-effects frequentist model across the full sample, with a very small p-value. That same analysis also flagged high between-study heterogeneity, meaning the individual studies varied considerably in how large (or small) the effect was — which the unsettled signals require be stated plainly. The meta-analytic data and code were made publicly available through Figshare, enabling independent verification.

• Dunne (2003) and Dunne (2005) report positive outcomes from PEAR laboratory remote perception trials spanning hundreds to millions of trials, with large z-values in the pooled analyses. Brenda J. Dunne developed the analytical and judging procedures — ternary coding and generalized descriptor systems — that made quantitative comparison across those experiments possible. Her spoke page Dunne: Remote Viewing Experiments and Analytical Methods covers this work in detail.

• Roe (2020), Mueller (2021), and Mueller (2017) each report positive hit rates in smaller-scale studies.

• Targ (2019) reports a positive result for the Pat Price formal series (9 trials), and Targ (1975) reports a positive result from the original 9-session Pat Price series using judges’ selection analysis.

• Brown (2012) reports positive clarity-score distributions across 11 targets, though no standardized effect size or p-value is present in the retrieved excerpts for that row.

• Escola-Gascon (2023) reports positive standardized effect sizes for believers/favorable-psi-attitude groups, with the largest effects in that subgroup. The overall sample effect (all 634 participants) was also positive.

Null results:

• Wahbeh (2021) found null overall performance on a Quick Remote Viewing task across 534 participants — the largest participant count among the individual studies in the table.

• Wiseman (2010) found null results across 4 blind-judging trials.

Below-chance (psi-missing):

• Wahbeh (2022) reports a below-chance Hedges’ g for extended perception (Quick Remote Viewing Task) pre-post change in a parent study result, with a hit rate of 0.19.

The coexistence of Wahbeh (2021) and Wahbeh (2022) as null/below-chance alongside the positive meta-analytic result from Tressoldi (2023) is an unsettled signal, not a resolved one.

Metric Breakdown (Incompatible — Not Pooled)

These metrics are not comparable and must not be averaged or pooled. Standardized effect sizes (Cohen’s d, z/√n), raw hit rates, prevalence proportions, and p-values alone each answer different questions about the data.

Unsettled Signals (Required Disclosure)

Three signals the evidence base itself flags:

1. Direction disagreement. Some studies are positive, Wahbeh (2021) and Wiseman (2010) are null, and Wahbeh (2022) is below chance. These are not reconciled by the current evidence.

2. High heterogeneity. At least one pooled result carries I² ≥ 50%, meaning the studies in that pool are not estimating a single uniform effect — individual study results vary more than sampling error alone would predict.

3. Subgroup inflation. The Escola-Gascon (2023) believers-group Cohen’s d (0.853) is substantially larger than the total-sample figure (0.457), raising the question of whether belief/attitude moderates the effect or whether the subgroup comparison is exploratory.

Methodological Context

The 2026 experts’ guidelines paper by Akin, Tressoldi, and Katz surveyed eleven experienced practitioners on recommended procedures for obtaining reliable remote viewing results — covering blinding, target selection, judging criteria, and session conditions. This represents an effort to codify best practices, though the guidelines paper itself notes that definitive empirical support for some of the recommended practices is still pending.

Marilyn Schlitz’s early replication work — moving the Targ-Puthoff protocol from SRI into independent laboratory settings, including transcontinental distances — is documented on her spoke page Remote Viewing Replication and Methodology.

Skeptical Critiques

What critics argue. The most persistent methodological objections to remote viewing research concern: (a) inadequate blinding in judging procedures, which could allow judges to match transcripts to targets using sensory cues rather than content; (b) the possibility that positive results concentrate in a small number of laboratories or viewers, making replication across fully independent sites the key unresolved question; and (c) publication bias — that null results are less likely to be submitted or accepted. The Katz-Tressoldi meta-analysis applied publication bias tests as part of its analysis, and the authors report no clear sign of publication bias in the retrieved page synopsis, though the specifics of which test and its result are not in the retrieved excerpts.

What the experimental data show. Wiseman (2010) — a null result across 4 blind-judging trials — and Wahbeh (2021) — null overall performance across 534 participants, the largest single-study participant count in the table — provide the clearest within-library evidence that positive effects do not always replicate under independent conditions. The high I² in the Tressoldi (2023) meta-analysis means the pooled effect size describes a heterogeneous distribution of outcomes, not a uniform one.

Analysis. Positive results appear across multiple independent laboratories and researchers over five decades. Null results also appear, including in large-N designs. The Escola-Gascon (2023) total-sample versus believers-subgroup gap (Cohen’s d 0.457 vs. 0.853) has not been resolved into a confirmed moderator versus a post-hoc artifact. Independent replication producing consistently positive results across laboratories with no pre-existing commitment to the protocol has not been demonstrated across the studies the library holds.

For the site’s curated synthesis, the Remote Viewing phenomena page and the Katz & Tressoldi meta-analysis spoke page provide the deepest integration of this evidence.