Exploratory record checks in 114 EEG and 87 fMRI adults gave positive heat rank covariance; three primary shared tests failed

We tested whether neural and report heat ordering covaried across people when shared trials were excluded from products. Public EEG and two fMRI studies supplied separate samples. Fixed readers used heat labels, not reports. The first test used 115 EEG and 20 MRI adults. Adjusted covariances were 0.007247 and 0.002305; the MRI interval crossed zero. Two shared claims failed. A later local lock preceded a sparse MRI extract and scores in 87 adults from another study. Its adjusted covariance was 0.001797, with 95% person Student interval [0.000112,0.003483]. Its shared Bonferroni p was 0.0554 across three named claims, so it also failed. All failures remain. Fixed descriptive checks then gave covariance 0.003812 across visits in 114 EEG adults and 0.003215 across runs in 87 MRI adults. Their pointwise intervals were [0.001548,0.006075] and [0.001550,0.004881]. An added exploratory audit covered 80 named one-sided human tests. The selected component Bonferroni p values were 0.0459 and 0.00945 under Student assumptions. This selected pair has no new independent confirmation. The first MRI run check remained uncertain. Exact trial exclusion, all people and omissions, model checks, source pins and terms, code, licensed extracts, figures and repeat steps are supplied. Source selection, whole-record fits, separate tasks and cohorts, sparse sampling, scanner and report effects limit scope. The result does not identify a pain-specific code, common neural cause or clinical biomarker.

pain discrimination · EEG · fMRI · AUC · measurement error · U statistics · person dependence · public data