At selected sites, sound groups concentrated repeat covariance in six SEEG patients and 24 fMRI adults; a within-group majority test failed

We tested a repeat covariance budget for 165 natural sounds in separate human studies. Three broad groups define two centered sound contrasts. One repeat set site scales; two other repeats scored group means and within-group differences. Development used two SEEG patients and four fMRI adults. A local code lock preceded tests in six other patients, 16 B2021 adults and eight NH2015 adults. Group mean differences held 35.87%, 38.79% and 34.38% of cohort aggregate repeat covariance. A fixed-size random group partition has expected share 2/164, or 1.22%. All 30 reserved people had positive excess group covariance. Mean excesses were 0.121993, 0.189140 and 0.125122. Their 95% person Student intervals were [0.029235,0.214750], [0.177788,0.200491] and [0.099084,0.151160]. The shared concentration test passed Holm correction for two claims, p=0.03125. The fixed within-group majority test failed, p=0.0625. One SEEG patient was negative and its cohort interval included zero. Spatial mean removal and 16-site panels retained positive concentration. Eleven finer labels reversed the majority in both MRI studies; electrical labels did not support that check. An exact projection reference and an unfitted reliability-screen model limit interpretation. Source selection can create positive repeat covariance from noise. These measurements do not establish latent neural information or a common cause. Clinical sampling, separate cohorts, common sounds, source screens and unequal measurement filters limit scope. The package preserves code, source pins and terms, plans, all results, figures, uncertainty draws and repeat steps. Original neural arrays stay external.

human neuroscience · SEEG · fMRI · auditory cortex · natural sounds · variance partition · repeat covariance · source selection