Exploratory vocal subgroup terms were positive for fixed sounds in six SEEG patients and 24 fMRI adults; shared tests failed

We tested whether auditory sites with a larger mean vocal preference have more relative vocal item spread. Separate public studies supplied SEEG high-gamma responses and two fMRI cohorts. Two patients and four MRI adults supplied development. A local code lock preceded new endpoints in six other patients and 24 other adults. One repeat and other sounds selected sites and scales. Two held-out repeats supplied within-class covariance. Mean pooled interactions were 0.1440, 0.2417 and 0.0976 in SEEG, B2021 and NH2015. Their 95% person Student intervals were [-0.2884,0.5765], [0.1993,0.2841] and [0.0075,0.1876]. After vocal subgroup means were removed, means were -0.0349, 0.0913 and 0.0037. The two fixed shared tests failed, with Holm p=0.8620 and 1.0000. All negative people and controls remain. An exact decomposition added after these results gave positive vocal subgroup-mean terms in all 30 people. Their mean terms were 0.1863, 0.1609 and 0.0965, with 95% person intervals [0.0439,0.3287], [0.1403,0.1815] and [0.0640,0.1289]. The SEEG person and sound refit range included zero, [-0.0479,0.3608]. This exploratory finding has no independent confirmation. Person and sound refits preserve uncertainty. An unfitted exact model uses one scalar feature and class mean shifts. A rectifier changes its interaction from zero to 31/12. Thus a positive interaction need not identify separate features for item differences in the two classes. A second calculation reproduced every primary endpoint, and fresh source audits checked all retained neural inputs. The package supplies code, source pins and terms, all results, figures and repeat steps. Original neural arrays remain external. This is a measured response benchmark and a failed test of cross-modal generalization. Sparse clinical coverage, source selection, separate cohorts, fixed sounds and unequal measurement filters limit scope. The test does not show an absent neural relation or a common cause.

auditory cortex · SEEG · fMRI · voice preference · repeat covariance · cross-modal test · negative result · rectification