The GML Project is a large-scale, multi-laboratory metascientific replication study investigating the claimed Ganzfeld telepathy effect — a purported form of anomalous information transfer between a sender and receiver under conditions of sensory shielding. Meta-analyses of over 100 existing Ganzfeld experiments have reported a statistically significant anomalous effect. The GML Project reinvestigates this claim using rigorous, preregistered confirmatory methods across multiple independent laboratories. Project A is the counterfactual "dry run": the full experimental procedure is executed without human participants, with all participant inputs replaced by a random event generator (REG). This serves as a metascientific control to verify the reliability and specificity of the experimental system before Project B — the participant-based study — begins.
The study protocol, including all data collection procedures, pre-processing rules, and statistical analysis scripts, was fully specified and locked prior to any data collection. The analysis code has been independently uploaded with a timestamp to: https://github.com/kekecsz/GML_project/blob/main/GML_Analysis%20script.R
Data collection status by site
| Laboratory | Status | Trials Collected | Verified |
|---|---|---|---|
| University of Freiburg | Complete | 252 | ✓ |
| Chapman University | Complete | 252 | ✓ |
| ELTE Budapest | Complete | 251 | ✓ |
| LMU Munich | Complete | 251 | ✓ |
| UC Santa Barbara | Complete | 201 | ✓ |
| IDOR Rio de Janeiro | Complete | 201 | ✓ |
| Total | Complete | 1,408 | ✓ |
Role-based access summary
All data access during the study was governed by Resarc's role-based access control system. Three role types were active across the study period:
No user held simultaneous write and audit privileges. All role assignments are logged and immutable.
Export log
| Event | Date | Performed by Role | Dataset Hash (SHA-256) |
|---|---|---|---|
| Full dataset export — Project A | 14 October 2025 | Principal Investigator | a3f82c...e7b14d |
| Auditor verification export | 16 October 2025 | Auditor | a3f82c...e7b14d |
Both exports produced identical checksums, confirming no data modification occurred between events.
Study lifecycle timeline
Note on analysis: Statistical analysis was conducted externally to the Resarc platform using the preregistered R script. The SHA-256 hash of the exported dataset is published above, allowing any independent researcher to verify that the analysed file is identical to the data as it left the platform. In-platform analysis verification is not yet available for this study.
Project A produced null results across all eight preregistered outcome measures (HR-X, HR-O, HR-SX, HR-SO, DR-X, DR-O, DR-SX, DR-SO), consistent with the expected behaviour of an unbiased experimental system in the absence of the independent variable. No alarming indication of system unreliability was detected. GML Project B — the participant-based study — has been cleared to commence.
The GML Project required end-to-end coordination across six independent international laboratories, with strict controls on data access, session execution, and export integrity. Resarc provided the platform infrastructure for the full data lifecycle — from session scheduling and role-based access control through to timestamped, tamper-evident data capture and auditable export.
For a study of this methodological rigour, data provenance is not a secondary concern — it is a primary one. The Resarc audit trail gives the GML team, participating labs, and any independent reviewer a verifiable record of exactly what happened to the data, at every stage, from the moment of collection.
Running a multi-site behavioural study that demands the same standard? Talk to us →