Debra Lynne Katz, PhD Sources:
Remote Viewing Meta-Analysis and Systematic Review
Debra Lynne Katz and Patrizio Tressoldi conducted the first comprehensive meta-analysis of all remote-viewing studies from 1974 through 2022, synthesizing five decades of research into a single quantitative assessment. Under their inclusion criteria, the systematic review reported remote viewing as yielding the largest pooled effect sizes among the ESP paradigms they compared. Whether this comparative advantage generalizes to the broader literature, or to operational reliability, remains a separate empirical question (see the Skeptical critique baseline section below).
Deeper dives — Katz:
Findings detail
- The meta-analysis identified a strong average effect size across all remote-viewing studies, indicating performance substantially above chance expectation.1
- Remote-viewing protocols demonstrated superior efficiency compared to other experimental approaches for testing extrasensory perception.1
- Analysis revealed minimal decline effect and no evidence of publication bias across the five-decade research corpus.1
- Both precognitive and clairvoyance tasks showed comparable effect sizes, with slight variations based on the presence of an outbound agent.1
Overview
Katz and Tressoldi’s 2023 systematic review represented the first comprehensive quantitative synthesis of remote-viewing research spanning nearly fifty years of investigation. The collaboration brought together Tressoldi’s expertise in meta-analytic methodology with Katz’s deep knowledge of remote-viewing practice and history as president of the International Remote Viewing Association. Their work addressed a fundamental question in parapsychology: how does remote viewing compare to other experimental approaches for studying extrasensory perception, and what does the cumulative evidence reveal about the phenomenon’s robustness.1
Historical context and research motivation
Remote viewing emerged from classified military intelligence programs during the Cold War, where it was used extensively by so-called “psychic spies” for intelligence gathering. The transition from classified to civilian research created a substantial but scattered literature base. By 2022, no comprehensive meta-analysis had synthesized this entire corpus. Katz and Tressoldi’s central objective was to discover how remote viewing differs from other ESP approaches in terms of experimental results and design components, and to identify central themes spanning five decades of investigation.1
The research team defined remote viewing as “the acquisition and description, by mental means, of information blocked from ordinary perception by distance or shielding and generally considered to be secure from such access.” This definition, drawn from foundational work in the field, encompasses both spatial separation (clairvoyance) and temporal displacement (precognition), distinguishing remote viewing from more constrained forced-choice paradigms.1
Key findings
The meta-analysis revealed a strong average effect size across all included studies, indicating that remote-viewing performance consistently exceeded chance expectation. Both frequentist and Bayesian analytical approaches converged on the same conclusion, strengthening confidence in the robustness of the findings.1
Effect size and hit rate statistics
After exclusion of outliers, the meta-analysis yielded an average effect size of .34 with a 95% confidence interval of .22 to .45. In terms of raw scores, these results correspond to a difference in hits score of 19.3% above expected chance, with 95% confidence intervals ranging from 13.6% to 25%. The analysis included 36 studies contributing 40 effect sizes. Both frequentist and Bayesian meta-analytic approaches produced consistent results, with the Bayesian analysis providing additional support for the robustness of the observed effects.1
The analysis found no evidence of publication bias, suggesting that the observed effects reflect genuine phenomena rather than selective reporting of positive results. Additionally, the research revealed only minimal decline effect across the decades of investigation, indicating that remote-viewing performance has remained relatively stable over time rather than diminishing as the novelty of the paradigm wore off.1
Publication bias and decline effect assessment
Publication bias assessment used standard funnel plot analysis and statistical tests to determine whether the published literature showed asymmetry indicative of missing negative results. The analysis found no significant evidence of such bias. The decline effect assessment examined whether effect sizes decreased systematically over the 48-year period from 1974 to 2022. While some ESP paradigms have shown substantial decline effects, the remote-viewing corpus demonstrated only minimal decline, suggesting that the phenomenon has proven relatively robust to repetition and increased scrutiny over time.1
Methodology and scope
The systematic review employed rigorous inclusion criteria to identify all relevant remote-viewing studies conducted between 1974 and December 2022. The research team developed operational definitions that distinguished remote viewing from other free-response ESP paradigms, ensuring that only studies meeting specific methodological criteria were included in the quantitative synthesis.1
Inclusion criteria and study selection
The inclusion criteria required studies to employ remote-viewing protocols as operationally defined by the research team, distinguishing these from other free-response approaches such as dream telepathy or ganzfeld studies. Studies had to report sufficient statistical information to calculate effect sizes. The search strategy covered both published journal articles and technical reports from the classified military programs that had been declassified and made publicly available. After applying inclusion criteria, 36 studies with a total of 40 effect sizes were selected for the meta-analysis. The team conducted both frequentist and Bayesian meta-analyses to ensure robustness of conclusions.1
The research team made all data and analytical code publicly available through Figshare, enabling independent verification and reanalysis of their findings. This commitment to transparency and open science practices strengthened the credibility of the meta-analysis within the broader scientific community.2
Moderator analyses
Katz and Tressoldi examined several potential moderator variables to determine whether specific design features influenced remote-viewing performance. The analysis compared precognitive tasks (where the target is selected after the viewing session) with clairvoyance tasks (where the target exists at the time of viewing), and examined the role of outbound agents who visit target locations.1
Precognition versus clairvoyance comparison
The moderator analysis revealed a small, nonstatistical difference between precognitive and clairvoyance tasks. Both types of tasks produced substantial effect sizes, but precognitive tasks showed slightly different patterns, particularly when compared to clairvoyance tasks with an outbound agent present at the target location. The presence of an outbound agent appeared to influence performance in clairvoyance conditions more than in precognitive conditions, though the differences did not reach statistical significance. This finding suggests that the temporal relationship between viewing and target selection may interact with other design features in subtle ways.1
The moderator analyses provided insights into which design features might optimize remote-viewing protocols for both research and practical applications. While the differences between conditions were generally small, they offered guidance for future protocol development and suggested avenues for more targeted investigation of specific design elements.1
Comparison with other ESP protocols
A central contribution of the meta-analysis was its comparison of remote-viewing results with effect sizes from other experimental approaches to studying extrasensory perception. This comparative analysis reported the largest effect sizes for remote-viewing protocols among the ESP paradigms in the inclusion set; the authors interpret this as a comparative effect-size advantage rather than a universal claim about RV’s evidential status.1
Cross-paradigm effect size comparison
The comparison examined meta-analytic results from ganzfeld studies, dream telepathy experiments, and forced-choice card-guessing paradigms. Remote-viewing protocols produced substantially larger average effect sizes than these alternative approaches. The effect size of .34 observed in remote viewing exceeded typical ganzfeld effect sizes and was markedly larger than forced-choice paradigm results. This superiority held even after controlling for methodological quality and publication bias. The finding suggests that the free-response nature of remote viewing, combined with its specific target selection and judging procedures, creates conditions particularly conducive to demonstrating extrasensory perception.1
The research team concluded that after more than fifty years of investigation into extrasensory perception, remote-viewing experimental protocols appear to be the most efficient for both experimental and practical applications. This conclusion has significant implications for resource allocation in parapsychological research and for the development of applied psi programs.1
Skeptical critique baseline
Remote-viewing research has accumulated a documented critical literature that any meta-analytic case must engage. The critiques below are recurring and well-cited in the parapsychology and skeptic-science literatures alike; the project’s response is to surface them — paired with serious replies — rather than evaluate the Tressoldi & Katz (2023) meta-analysis in isolation from the field’s adversarial history.
Critique 1: Critique 1: Sensory cues in early SRI remote-viewing transcripts
Skeptic source: Marks & Kammann (1978, 1981) re-judged transcripts from the foundational Targ–Puthoff SRI remote-viewing studies after stripping extraneous cues (target descriptions, weather references, distinctive landmarks). Independent judges could no longer match transcripts to targets above chance. The 1980 book The Psychology of the Psychic (Prometheus Books) expanded the cue-discovery argument across the early RV literature.[4]
Rebuttal: Post-SRI remote-viewing protocols (1980s onward, including most studies in the Tressoldi–Katz inclusion set) adopted double-blind judging, randomized target pools, and standardized transcripts specifically to close the cue-discovery objection. The pooled effect sizes reported in the 2023 meta-analysis derive from protocols designed to be cue-immune; the Marks–Kammann critique applies primarily to the historical foundational studies, not to the contemporary corpus.[1]
Analysis. — protocol generations matter: the original SRI critique is well-documented and widely accepted; whether modern protocols fully close it remains a separate empirical question that requires explicit adversarial replication.
Critique 2: Critique 2: Judging-protocol and free-response statistical objections
Skeptic source: Hyman (1985), in his critical appraisal of the ganzfeld psi experiment in the Journal of Parapsychology, raised methodological objections — selective reporting, multiple analyses, judging variability, inadequate randomization — that the skeptical and meta-scientific literatures have treated as foundational for all free-response psi paradigms, including remote viewing.[5]
Rebuttal: The 1986 Hyman–Honorton Joint Communiqué (Journal of Parapsychology, 50, 351–364) accepted that “there is an overall significant effect that cannot be reasonably explained by selective reporting or multiple analysis,” while Hyman and Honorton continued to disagree on the psi interpretation. The Communiqué specified methodological standards that have shaped subsequent ganzfeld and remote-viewing protocols, and the Tressoldi & Katz (2023) meta-analysis applied 3-Parameter Selection Models and Robust Bayesian Meta-Analysis specifically to test for publication bias and selective reporting in the modern corpus.[1]
Analysis. — Hyman and proponent statisticians (notably Jessica Utts) remained divided in the 1995 AIR evaluation on whether modern statistical controls fully satisfy the original critique.
Critique 3: Critique 3: Laboratory effects do not translate to actionable operational intelligence
Skeptic source: Mumford, Rose, & Goslin (1995), in the American Institutes for Research / CIA evaluation of the 20-year Stargate program, concluded: “remote viewings have never provided an adequate basis for ‘actionable’ intelligence operations — that is, information sufficiently valuable or compelling so that action was taken as a result,” and noted “a large amount of irrelevant, erroneous information is provided and little agreement is observed among viewers’ reports.” The CIA terminated the Stargate program in 1995 based on this evaluation.[6]
Rebuttal: The AIR/CIA panel’s statistician member, Jessica Utts, concluded that the laboratory data showed statistically significant effects (5–15% above chance for some subjects). The Mumford–Rose–Goslin conclusion was specifically about operational/actionable utility rather than laboratory significance — they recommended “a higher level of critical research and tighter controls” rather than dismissing the underlying effect. The Katz-associated Project Firefly applied-ARV trading project (see the ARV applied-forecasting spoke) is one transparent contemporary example of an operational ARV effort that failed precisely the operational-utility test, demonstrating that the AIR critique remains live in current practice.
Analysis. — the operational-actionability gap between laboratory remote-viewing effects and field-intelligence reliability is consistently observed; the Tressoldi & Katz (2023) meta-analysis addresses laboratory effects, not operational reliability.
Implications for research and practice
Within the meta-analytic models applied, the authors reported a positive pooled effect that did not decline materially across decades. Whether this supports operational utility — as distinct from laboratory-level statistical effect — depends on independent high-quality replications, preregistered protocols, and adjudication of judging and file-drawer artifacts. The Skeptical critique baseline section below summarizes the historical objections that any modern remote-viewing case must engage.1
Recommendations for future research
Katz and Tressoldi identified several directions for future investigation. They recommended more systematic exploration of moderator variables, including target characteristics, viewer training levels, and judging methodologies. The minimal decline effect observed suggests that remote viewing may be particularly suitable for long-term research programs and practical applications. The research team also emphasized the need for continued methodological refinement, particularly in developing standardized protocols that balance experimental rigor with the flexibility required for optimal viewer performance. Their work with the International Remote Viewing Association’s Research Unit reflects this commitment to advancing both the science and practice of remote viewing.13
For practitioners and organizations interested in applied remote viewing, the reported pooled effect sizes are larger than ESP-paradigm averages in the inclusion set, and minimal decline was observed. Whether these laboratory effects translate to operational reliability is a separate question; the 1995 AIR/CIA evaluation of the Stargate program concluded that remote-viewing data, even when statistically present in lab conditions, did not produce actionable intelligence (Mumford, Rose, & Goslin, 1995). The Skeptical critique baseline section below details the historical record on this gap.1
References
- Tressoldi, P. E., & Katz, D. L. (2023). Remote Viewing: A 1974-2022 Systematic Review and Meta-Analysis. [preprint] Journal of Scientific Exploration, 37(3), 467-489. [Tressoldi 2023] R001 ↩︎
- Tressoldi, P., & Katz, D. (2023). Remote Viewing Meta-Analysis. Figshare. [Tressoldi 2023] R002 ↩︎
- Katz, D. L., & Tressoldi, P. (2021). Remote viewing applications: An historical overview and a new survey. [ResearchGate] [Katz 2021] R003 ↩︎
- Marks, D., & Kammann, R. (1978). Information transmission in remote viewing experiments. Nature, 274, 680–681. [Nature]; Marks, D. (1981). Sensory cues invalidate remote viewing experiments. Nature, 292, 177. [Nature]; Marks, D., & Kammann, R. (1980). The Psychology of the Psychic. Prometheus Books. [Marks 1978] R004 ↩︎
- Hyman, R. (1985). The ganzfeld psi experiment: A critical appraisal. Journal of Parapsychology, 49, 3–49. [subscription required: Journal of Parapsychology]; Hyman, R., & Honorton, C. (1986). A joint communiqué: The psi ganzfeld controversy. Journal of Parapsychology, 50, 351–364. [ResearchGate] [Hyman 1985] R005 ↩︎
- Mumford, M. D., Rose, A. M., & Goslin, D. A. (1995). An Evaluation of Remote Viewing: Research and Applications. American Institutes for Research, prepared for the Central Intelligence Agency. [National Security Archive PDF] [Mumford 1995] R006 ↩︎
Deeper dives — Katz: