A shared sound-order test failed in 20 iEEG patients and 16 fMRI adults; common-average reference can create a perfect reversal from one monotone drive

We tested whether two auditory sites change their response-order relation between human vocal and other sounds. Separate public studies supplied signed intracranial EEG and BOLD fMRI. Four people per modality supplied development. A fixed test used 20 other patients and 16 other adults. Calibration sounds selected one site pair per person. Other sounds tested that pair. The score was minus the product of two Kendall tau-a values. Mean scores were -0.00849 and +0.03817. The 95% person Student intervals were [-0.02517, +0.00819] and [+0.00982, +0.06653]. Positive scores occurred in 8/20 and 11/16 people. The shared positive test failed, p=0.850038. Crossed person and test-sound ranges included zero in both modalities. Swapped sound halves, small candidate sets, and single MRI estimates weakened the MRI result. An exact model gives a separate method limit. Three native responses all increase with one scalar. Common-average reference creates a perfect rank reversal, with score one, in separate calibration and test samples. No noise or human fitting is needed. Thus, stable reversal after reference does not count native neural drives. The human tests do not establish a positive shared neural effect or a cause. Cohort differences, clinical sampling, source processing, unequal site counts, and correlated errors limit scope. The package preserves code, permitted electrical inputs, source versions and terms, all outcomes, figures, plans, and repeat instructions. Original author arrays with unconfirmed reuse terms remain external.

human neuroscience · intracranial EEG · fMRI · auditory cortex · rank association · common-average reference · measurement model · negative result · cross-modal test