A fixed -10% hand response bound passed in 14 ECoG patients and 58 fMRI adults under Student mean assumptions

We tested hand and tongue responses at sites selected with separate data. Public human studies supplied ECoG and BOLD fMRI. A local rule fixed four shared claims. Input audits then corrected MRI development identities and excluded one repeated ECoG record. Corrected confirmation used 14 patients and 58 adults. At tongue-selected sites, mean N+0.1P was 15.003 electrical power points and 2.161 MRI source units. Their 95% person Student limits were [5.424,24.581] and [0.954,3.368]. P was positive in each study. The shared four-claim Holm p was 0.009788. Thus mean hand responses exceeded -10% of mean tongue responses under this assay and mean model. The reverse bound failed, Holm p=0.111181. Both shared decrease tests failed. The passing electrical bound interval crossed zero with median reference or preceding rest. This is not processing-independent neural generalization or evidence of absent inhibition. An audit found 18 exact repeated movement/rest pairs in one patient. A tested version URL also returned other subject IDs. All 124 final MRI files match version-tagged hashes. These are disclosed, corrected tests in reused public people, not an unchanged external preregistration. Cohorts, movements, coverage, baselines and signal filters differ. The package gives original code, source pins, nominal and corrected results, figures and repeat steps. Original neural arrays remain external.

human neuroscience · ECoG · fMRI · motor cortex · tongue · hand movement · negative BOLD · independent data · data provenance · exact duplication · response bound