The original correlation-matrix experiments conducted by Lucadou's group (1986–2006) reported significantly more significant correlations between psychological and physical variables in experimental sessions than in control sessions.

The reports exist and were computed in good faith, but the significance counting assumed independent cells in matrices of intercorrelated variables, and the program's own later statisticians report that dependence-aware simulations diminish the results. Until MPI-R002 is run, the founding surplus is neither restored nor dissolved.

The literature record clearly demonstrates that Lucadou's group and Walach's 2020 replication reported significantly more significant correlations than expected by chance. Whether these reported effects are artifacts of statistical dependence is contested, but the fact that these reports form the historical foundation of the program is established. The strongest opposing consideration is that the raw data remain unavailable for independent verification, but the existence of the published reports is undisputed.

The existence and qualitative direction of the founding reports are well documented: Lucadou’s group and Walach’s 2020 paper clearly state that experimental sessions yielded more significant psychological–physical correlations than chance or control matrices (MPI‑E001, MPI‑E002). What is in doubt is the strength and anomaly‑content of those surpluses, because Flores’ later work shows that psychological variables were highly intercorrelated and that dependence‑aware simulations reduce the apparent excess (MPI‑E007, MPI‑E009). The strongest opposing consideration is precisely that dependence critique: if the original analyses substantially over‑counted significance by ignoring clustered structure, then the “far more than chance” characterization is overstated, and in the absence of a full reanalysis of the 1986–2006 raw data we cannot upgrade this claim beyond provisional support.

Multiple successor sources in this file, including a verbatim Walach et al. 2021 characterization and the 2020 two-lab paper, consistently describe Lucadou’s 1986–2006 series as reporting more significant experimental than control correlations. The claim as stated is about those reports, not about whether the counting was statistically valid, and on that narrower reading the record is unambiguous. The strongest opposing consideration is that the original full texts and raw data were not examined here, so the characterization is second-hand, and the dependence critique in MPI-C009 attacks the very surplus those reports claimed.

Assessed as stated, this is a report-level claim: that Lucadou's group published analyses showing surpluses of significant matrix cells in experimental versus control sessions. That is multiply attested inside the corpus, including by the Walach et al. 2021 introduction and Grote 2021, and even the program's sharpest internal critic (Flores) accepts the original analyses reported these numbers while disputing their validity. The strongest opposing consideration is provenance: the founding full texts were not independently examined for this case and the raw data were never published, so the characterization rests on later descriptions by proponents and successors. The deeper attack — that the surpluses are artifacts of significance counting under violated independence — is real, but it is the business of MPI-C009, not of this claim, which asserts only what was reported.

That the founding reports exist and assert an excess of significant cells relative to chance and control matrices is attested verbatim in successor papers by Walach and by Grote, so the documentary content is secure. The word "significantly," however, presupposes a valid null distribution, and the analyses counted cells against an independence assumption that matrices of intercorrelated behavioural variables almost certainly violate. The strongest opposing consideration is auditability: the primary 1986–2006 texts were not examined for this corpus, the early series is German-language and pre-preregistration, and the raw data have never been published, so no independent check of the surplus is currently possible. I therefore credit the report but not the inference it encodes.

Nothing in the new material touches the founding datasets. The reports exist and were computed in good faith, but the significance counting assumed independent cells among intercorrelated variables, and the program's own later work (MPI-E007) says dependence-aware simulations diminish the results. Unresolved until the founding data are reanalyzed.

The reports are credible as reports, but the claim asserts a significant surplus, and that significance counting assumed independence among intercorrelated cells — a critique raised from inside the program (MPI-E007). The raw data remain unpublished, so the founding result is still unauditable.

The reports exist and were computed in good faith, but the significance counting presumed independent cells among intercorrelated variables, and the program's own dependence-aware simulations (MPI-E007) diminish the founding results. Nothing in this run touches the original datasets, and raw data remain unpublished.

The reports were made and are credible as reports, but the claim asserts a significant surplus, and the significance counting assumed independence among intercorrelated matrix cells — a critique originating inside the program (MPI-E007). Raw data remain unpublished (MPI-E001), so the founding result stays unauditable.

The existence of these positive reports in the founding literature is confirmed by later reviews and replications (MPI-E001, MPI-E002). However, because the original datasets utilized statistical methods vulnerable to the dependence critique, the substantive finding of anomalous correlation remains provisionally supported rather than definitively established. The strongest opposing consideration is that dependence-aware reanalyses strongly suggest the initial statistical significance was an artifact of intercorrelated psychological variables.

Multiple later sources, including Walach et al. 2020/2021 and Grote 2021, report that Lucadou’s original matrix experiments showed more significant psychological–physical correlations in experimental than control sessions, and there is no evidence in this corpus that this characterization is factually wrong (MPI-E001, MPI-E002). The main doubt concerns not whether those surpluses were reported, but whether they were computed against a valid null distribution, given later dependence-aware critiques (MPI-C009). The strongest opposing consideration is Flores’s finding that high intercorrelations among psychological variables inflated early significance counts, which suggests the original “far more than chance” framing overstated the anomaly even if the basic reporting description is accurate.

Successor papers (Walach 2020/2021, Grote 2021) consistently describe Lucadou's 1986–2006 series as reporting more significant psychological–physical cells in experimental than control matrices, and Walach 2020 itself added a further surplus against 10,000 simulated experiments. The claim as stated is about what those experiments reported, not about whether the counting was valid, and that reporting fact is multiply attested. The strongest opposing consideration is that the original full texts and raw data were not independently reanalyzed here, so the numerical surpluses remain second-hand, and the dependence critique in MPI-C009 attacks their statistical meaning rather than their existence as reports.

That the founding reports exist and assert an excess of significant cells relative to chance and to control matrices is directly attested by the proponents' own later summary (Walach 2021) and is not in dispute. What is weak is the evidential status of the assertion: the primary texts were not examined in this corpus (MPI-E001 explicitly relies on successor-paper descriptions), the 1986–1987 series is largely German-language and second-hand, and the raw data have never been released for reanalysis. The strongest opposing consideration is MPI-C009: the word "significantly" in this claim presupposes a null distribution that intercorrelated psychological variables violate, so the reported surpluses may be a counting artifact rather than a measured excess. Verdict is therefore secure for the reporting and provisional for the statistic.

The reports exist and were made in good faith, but the claim asserts a *significant* surplus, and the significance counting assumed independent cells among intercorrelated variables — a critique originating inside the program (MPI-E007). Raw data remain unpublished, so the founding result is unauditable. Gemini's reading as a pure reportage claim understates what the word 'significantly' commits to.

The reports are genuine, but the claim asserts a significant surplus, and that significance count assumed independence among intercorrelated matrix cells — a critique originating inside the program (MPI-E007). Raw data remain unpublished (MPI-E001), so the founding result is unauditable.

That these reports exist and say what they say is well documented. That the reported surpluses were computed against valid null distributions is exactly what the dependence critique (MPI-C009) contests: the original significance counting assumed independent cells in matrices built from intercorrelated psychological variables.