Fixed tests did not confirm a shared positive memory relation or EEG–fMRI transfer of repeatability for 102 videos

We tested whether independent video memory scores predict response-pattern repeatability and whether repeatability profiles transfer between measurements. Separate public studies supplied six scalp EEG adults and ten fMRI adults viewing 102 common videos. Selected content controls and response size from disjoint session halves entered fixed rank tests. Mean adjusted memory correlations were +0.049 and +0.065, with person 95% intervals [-0.054, +0.152] and [+0.023, +0.105]. Mean correlations with the other modality's observed profile were +0.004 and -0.007, with intervals [-0.136, +0.144] and [-0.074, +0.060]. Neither fixed shared claim passed. Shared Holm p values were 0.25 and 0.591797. Crossed resampling rebuilt both donor maps. An added precision audit found negative adjusted half-profile agreement in nine MRI people. A standard scalar identity gives residual correlation minus the original correlation under mutual linear control; 256 rational cases checked it. This is an interpretation limit, not a fitted human cause. A source audit recovered column order in 80 MRI region files and checked 51,960 values. Small moment files preserve the assay. Separate cohorts, audio and task differences, sparse EEG coverage, imperfect controls, and source processing limit scope. No causal or shared-code claim is established. The package gives source pins, terms, code, all results, uncertainty draws, figures, and repeat steps. Original media and arrays remain external.

human EEG · fMRI · video memorability · pattern repeatability · cross-modal validation · source audit · reproducibility