Voice group axes had more covariance per axis in human SEEG and fMRI; the fixed majority test failed

We tested how three voice groups divide repeat covariance for 165 natural sounds. Separate public studies supplied SEEG high-gamma amplitude and BOLD fMRI. Two patients and four MRI adults supplied development. A local code lock preceded tests in six other patients and 16 other adults. Within-group mean covariance fractions were 64.13% and 59.50%. The SEEG majority margin was 0.098904, with 95% person Student limits [-0.034409,0.232217]. The fixed full shared rule failed, p=0.114814. One negative SEEG patient remains. A smaller planned component compared covariance per axis. Each voice group axis carried more covariance than an average within-group axis. Ratios of cohort mean terms were 45.30 and 55.13. A later Holm audit included all six locked primary components. Its shared adjusted p for this smaller result was 0.046875. This reduced-claim audit was added after the full test failed. It has no new independent electrical acquisition confirmation. Eight records from another MRI study gave a 66.76% within-group fraction. All controls, sound resampling and negative results remain. A repeat-shared artifact can also give the pattern with class-constant neural responses. Clinical sampling, source screens, a common retained electrical order, different cohorts, tasks, filters and site weights limit scope. This is a benchmark for processed repeat covariance, not a common neural cause or an information fraction. The package gives code, source versions and terms, plans, all results, figures and repeat steps. Original arrays remain external.

human neuroscience · SEEG · fMRI · voice groups · repeat covariance · within-group variation · orthogonal projection · negative result