A User Facility for RV Research

Institut Métapsychique International

Paris, France

1. Overview

Project Goal: To Create a User Facility for RV Science

The Institut Métapsychique International (IMI) will collaborate with Social-RV to create a research gateway for Remote Viewing (RV). Rather than a single study, this project builds a lasting scientific infrastructure—a Portal through which independent researchers can design, preregister, and execute experiments at scale, transparently, and with the highest standard of protocol verifiability. The goal is a user facility that attracts scientists across disciplines to undertake research in this controversial and potentially ground-breaking field. The Portal allows experiments to be done faster, with less overhead and risk, and with a degree of verification that is unprecedented. Ultimately, science is collaborative and cumulative: this is our approach to leveraging Social-RV’s technology.

To create the Portal, we will combine IMI’s expertise in large-scale psi data modeling and online experimentation with Social-RV’s advanced infrastructure and AI-judging system.

This proposal is distinctive in that the Portal integrates three features never before combined in psi research: (1) cryptographic protocol assurance via blockchain, (2) AI-assisted scoring of free-response RV data, and (3) a standardized, cross-institutional user facility enabling parallel, high-throughput experimentation.

Stages of the Project: Calibrate, Build and Test.

Calibrate Social-RV’s AI-based judging. Test the 4500+ RV sessions for a RV effect.
Build the Portal infrastructure in collaboration with the Social-RV team.
Live-test stages 1 and 2 with two pilot studies.

Timeline: 12 months

Completion of stages 1 and 2: at 6 months.
Completion of stage 3: at 12 months.

2. Introduction

Background and Motivation

Remote Viewing (along with the concomitant Ganzfeld protocol) has consistently produced evidence for nonlocal perception since the 1970s (Baptista et al., 2015). These “free-response” protocols supply the best scientific evidence for psi effects of nonlocal perception. Yet limitations of the protocol impede our understanding, as well as any acceptance by mainstream science.

Three key limitations to RV research are:

Labor-intensive, non-standardized analysis (judging).
The resultant small sample sizes that limit power and reproducibility (Tressoldi, 2023).
Skeptical criticism of protocols.

The Social-RV platform solves these obstacles by providing blockchain authentication and AI- judging of RV sessions. However, these are not yet sufficient to do real science because:

Valid scientific results require well-calibrated analytical methods.
Calibrated procedures are needed to test existing data and estimate effect sizes.
50 years of RV makes clear that consensus and knowledge will advance only through a
visible program of multi-disciplinary research. This requires an open-access infrastructure to attract research across fields: a user facility.

Most importantly, preliminary summaries from social-rv.com indicate a strong RV effect (see Methods – Track 2). While highly encouraging, these findings underscore the need for rigorous data verification and the establishment of a solid foundation for analysis and communication. Confirming these early results would make the proposal even more compelling: a user facility would drive rigorous science by providing a natural home for the necessary independent replications.

The Concept of a User Facility.

In physical sciences, user facilities—telescope arrays, particle accelerators—accelerate discovery by providing open-access, calibrated tools for shared investigation. We propose the same model for psi science: the RV Portal as a user facility where researchers can access the resources and participant pool of Social-RV, augmented by rigorous calibration of AI scoring systems. The Portal supports a research program: it provides for parallel experiments, innovative experimental designs and system upgrades.

IMI’s Psi@Home initiative (Bancel et al., 2025) has explored the feasibility of this model with forced-choice psi tasks. Extending it to free-response RV through Social-RV infrastructure will establish the world’s first open-access observatory for nonlocal perception.

Scientific Impacts.

No doubt, the headline impact will be whether an RV effect emerges from the data. However, we emphasize the long-term impacts that are central to our proposal and will emerge organically:

The first authenticated analysis of a large AI-judged RV database: a new standard.
A ground-breaking facility for gold-standard replications open to the scientific community.
Acceleration via new experimental designs that are fast, registered and validated.
Standardized replications for education and undergraduate projects.
Building a community of multi-disciplinary users exploring nonlocal perception.

The project establishes the methodological and procedural foundation for RV research at scale – turning an exploratory practice into solid science. Regardless of effect outcomes, it creates the infrastructure for open, transparent, and cumulative progress in consciousness research.

Vision.

While the immediate focus is on the rigorous validation and calibration of Social-RV's existing database and methodologies, the Portal serves as a durable infrastructure for future academic partners. The Portal is designed to evolve into the world's most accessible platform for RV research, enabling studies that address long-standing questions currently hampered by technical hurdles. For instance, the Portal will allow for large-scale performance profiling (e.g., correlating RV success with psychological metrics like cognitive flexibility or tolerance for ambiguity); target optimization studies (testing if AI-quantified image properties, such as visual complexity, predict session success); longitudinal learning curves (comparing performance progression between different training cohorts over multiple years); and studies that associate the cognitive impacts of RV as a practice. By creating the Portal, we transform Social-RV into a continuously operating research laboratory capable of deploying novel, high-impact experiments.

While some may be concerned that standardized protocols can reduce effect sizes, the Portal is explicitly designed to study such questions empirically, enabling comparisons across protocol strictness, feedback delays, task types, and viewer populations.

3. Methods

Track 1 – Calibrate and assess AI-judging.

Goal: Determine the behavior of AI-judging under true-null and simulated-effect conditions.

Create surrogate datasets by permutations of decoys while preserving session structure. Original target images are replaced by decoys.
Run AI-judging simulations on surrogates to estimate the null distribution of hit-rates and ranks and the variances. Simulations will run until the null distribution is stable (no material chance with additional iterations).
Produce null distribution curves and a model of chance expectations. Because AI-judging is new and priors are impractical, use a frequentist paradigm with Fisherian discovery (as opposed to Bayesian analysis).
Model simulated effects (e.g., estimate ES ≈ 0.15 (Utts, 1999; Baptista et al., 2015)) to assess power and sensitivity. Model development will be necessary for this stage.
Assess and compare stability and power for different numbers of decoys (10 → 6 → 4).
Compare with simulations using drawings/sketches and written inputs separately.
Prepare exploratory analysis procedure for testing orthogonality/similarity of targets.
Deliverable → AI-Judging Calibration Report and code archive. Analysis of power versus decoy number.

Track 2 – Dataset analysis.

Preliminary Results: Social-RV posts summary statistics on its website. AI rank scores of 4,526 RV sessions give a Spearman rank p-value = .014 (ranks 1 to 10 with 1 being the best match). We have verified Social-RV’s posted one-tailed p-value. However, the test choice is a poor one. A binomial test is far more powerful, yielding p<.000000001 (sic). Furthermore, the effect size per session is ES~.09, which aligns well with published effect sizes for ganzfeld and RV protocols (Utts, 1999; Baptista et al., 2015; Tressoldi et al., 2023). These encouraging results highlight the importance of validation.

Goal: Apply the calibrated metrics to the full Social-RV dataset. The use of careful data splitting is crucially important. Given the preliminary results, data-splitting can be conservative, starting with 1/4 of the full dataset, or less. Re-tests will be pre-stated and registered.

The following sketches the main aspects of analysis. Precise analyses will depend on results of track 1.

Primary analyses: Exact binomial tests of rank scores, mean rank and hit-rate (rank 1); test position effects across multiple user trials. Power analysis simulations to complement tests.
Approach: exploratory frequentist modeling with permutation tests; sequester pre-split dataset for confirmatory vs. exploratory phases to avoid data-mining pitfalls.
Outputs: Null distributions for primary analyses, power estimates, and effect-size benchmarks for each tasking type (precognitive-blockchain, weekly, practice, etc.).
Deliverable → Open dataset + analysis notebook + paper to peer-reviewed journal.

Portal Track – Build the Portal.

Goal: Upgrade Social-RV from a website for RV practitioners to include a user facility for researchers. The Portal is designed by the IMI team and implemented by Social-RV engineers, per the RFP.

Risk mitigation: Portal development is contingent on assessment of the AI-judging system. Barring the discovery of serious technical flaws or obstacles, development can proceed even if a null result is found for the RV effect. In the event of serious technical issues, portal development will proceed only after their resolution by the Social-RV team.

The Portal includes the following functionalities:

Registration of research projects and verification of acceptance criteria.
Accounts for Principal Investigators; dedicated repositories for data.
Public pages for each project to invite participants and communicate progress.
Access to AI-judging and blockchain verification. User manuals for researchers.
Consent to a Portal charter of conduct and responsibilities (preregistration; open data access; possible fee option, etc)

Modalities and governance procedures will be developed in collaboration with Social-RV. Some aspects may use third-party resources (ResearchHub, GitHub, OSF etc.). The guiding principles will be transparency and simplicity for researchers.

Track 3 – Test the Portal

Goal: Two preregistered pilot experiments will test the Portal’s functionality. Each study will recruit ~35 participants (≈ RV 180 trials).

Track 3 tests Portal functionality using Social-RV’s current AI-judging system and infrastructure. The sample size (N≈35 per group, ~180 trials) is chosen to validate workflows, effect-direction, and not to formally test an RV hypothesis or make effect-size estimates.

Pilot A: Social-RV active users.
Pilot B: Psi@Home cohort, allowing comparison between communities.
Participants complete consent forms and profile surveys.
Participants contribute 4 - 6 of blockchained RV sessions using the current AI-judging system.
First sessions are preceded by a brief introduction to RV and guided tryout session. Participants have ~1 month to complete sessions
Data remain sequestered until both studies close.
Analysis: Exact binomial test of rank-split judging scores (ranks 1-5) against null expectation and ranks 6-10. The exploratory tests will be pre-registered and the exact statement will depend on the assessment of Track 2.
Deliverable → Demonstration of portal functionality + participant feedback; peer- reviewed publication.

Protocol elements for the pilot studies:

Short video for uniform introduction to remote-viewing sessions for all participants.
Personal video meetings to clarify the procedure, answer questions, and encourage motivation.
Participants provide email and/or cell phone contacts for communication with the team.

Feasibility: IMI has conducted two studies in this format for the Psi@Home Project (Bancel, et al., 2025; Bancel, 2025), where ~60 participants completed >200 15-minute sessions of a forced-choice task within 3 months. Note that typical RV sessions often last about 15 minutes. Given the existent participant pools at Psi@Home and Social-RV, the 6 month time budget for completion of the pilot studies is safely conservative.

4. Institution and Team

The Institut Métapsychique International (IMI), founded in 1919 in Paris, France, is a non-profit research institute dedicated to scientific study of consciousness and psi phenomena. Recent work includes the Selfield series, the Psi@Home platform, analysis of large parapsychological databases (PEAR, GCP), and numerous peer-reviewed publications on experimental methodology. The IMI is a recipient of the Parapsychological Association’s 2025 Book Award (Evrard et al., 2025).

Principal Investigator: Peter Bancel, PhD. Experimental Physics

Co-Investigators / Collaborators:

Mario Varvoglis, PhD. Experimental Psychology: Portal design and project consultant.
Jérémie Phillippe Masters Physics; Philosophy: Statistical and analysis support; pilot study.
Anaïs Bensarha: BSc Philosophy: Pilot study manager.

5. Budget : 52k€ ($60kUS)

Role	Name	Time	Salary	Responsibilities
PI	Bancel	0.4 FTE	38k€	All project aspects
Co-I	Varvoglis	0.1 FTE	Volunteer	Portal design, protocols
Data/Stats	Phillippe	0.2 FTE	Volunteer	Simulations / analysis
Pilot Coordinator	Bensarha	0.3 FTE	8k€	Participant manager
Engineering (Social-RV)	TBD	In-kind	Social-RV (offered)	Judging system Portal backend/frontend
IMI	Admin	Overhead	6k€ (~12%)	Institutional support

6. References

Bancel, P.A., Varvoglis, M., Boban, J., & Bensahra, A. (2025). A Forced-choice Precognition Experiment with Selected Cohorts. Journal of Anomalous Experience and Cognition, 5(1), 14-46.

Bancel, P. A., (2025). KPU Registered Study 1092: Summary of Results.https://www.koestler-parapsychology.psy.ed.ac.uk/Documents/KPU_1092_Results_Summary.pdf

Bancel, P. A. (2017). Searching for global consciousness: A 17-year exploration. Explore, 13(2), 94-101.

Baptista, J., Derakhshani, M., & Tressoldi, P. E. (2015). Explicit anomalous cognition: A review of the best evidence in ganzfeld, forced choice, remote viewing and dream studies. Parapsychology: A handbook for the 21st century, 192-214.

Evrard, R., Berghmans, C., & Rabeyron, P. L. (2025). Grand manuel de parapsychologie scientifique. Dunod.

Storm, L., & Tressoldi, P. (2023). Assessing 36 Years of the Forced Choice Design in Extra Sensory Perception Research: A Meta-Analysis, 1987 to 2022. Journal of Scientific Exploration, 37(3).

Tressoldi, P. (2023). Remote Viewing: A 1974-2022 Systematic Review and Meta-Analysis. Journal of Scientific Exploration, 37(3).

Utts, J. (1999). The significance of statistics in mind-matter research. Journal of Scientific exploration, 13(4), 615-638.

Varvoglis, M., & Bancel, P. A. (2015). Micro-psychokinesis. Parapsychology: A handbook for the 21st century, 266-281.

Appendix – Methodological Precedents & Analytical Expertise

A.1 Background

The proposed calibration and analysis will draw on the IMI team’s expertise in modelling and statistical analysis of psi data.  IMI researchers have contributed influential re-analyses of complex parapsychological databases.  These studies collectively demonstrate the team’s capability to handle non-standard, high-dimensional data with rigorous statistical controls, simulation modelling, and open-data reporting.

A.2 Key Precedents

PEAR Laboratory Micro-PK Database Re-analyses

Output: Elitist hypothesis favored over PEAR’s universalist interpretation.
Contributors: P. Bancel (lead analysis), M. Varvoglis. (Varvoglis & Bancel, 2015).

Consortium Replication of the PEAR Data: Re-assessment

Output: Showed Consortium’s null result due to incorrect power analysis.
Contributors: P. Bancel (lead analysis), M. Varvoglis. (Varvoglis & Bancel, 2015).

Micro-PK meta-analysis reappraisal

Output: Heterogeneity of studies refutes publication bias hypothesis.
Contributors: P. Bancel

Global Consciousness Project (GCP) Analyses

Output: Refutation of GCP hypothesis.
Contributors: P. Bancel (Bancel, 2017).

Psi@Home Experiments (2022 – 2025)

IMI’s own online platform for forced-choice psi tasks.
Demonstrated feasibility of remote participation, preregistration, and modular experimental design.
Produced significant per-session effect sizes (0.10 – 0.20) under fully blinded conditions.
Design and analysis: P. Bancel & M. Varvoglis (co-developers), A. Bensarha, J. Philippe (Bancel et al., 2025).

A User Facility for RV Research

Proposal

Proposal

Proposal

Proposal

A User Facility for RV Research

Proposal

Proposal

Funding Progress

Nonprofit Recipient

Peer Reviews

Topics

DOI

Funding Progress

Nonprofit Recipient

Peer Reviews

Topics

DOI