An 8 s word repetition term did not pass auditory response tests in six iEEG patients and 25 fMRI listeners

We tested whether recent word repetition adds a shared suppression signal during story listening. A future repetition trace provided a control for symmetric exposure. We reused public auditory responses and sound features. Three iEEG patients selected an 8 s interval. We fixed it before tests in six other patients and 25 fMRI listeners from another study. Each person had a separate model. The baseline included sound, word timing and simple word features. Mean test R-squared gain was -0.00004972 in iEEG, with a 95% person interval from -0.00013954 to +0.00003455. The fMRI gain was -0.00038189, with an interval from -0.00110136 to +0.00031262. Both primary Holm p values were 1. The fixed shared rule failed. All planned sensitivity variants had negative mean gains. An added unconstrained diagnostic also had a negative mean in each modality. This result does not establish that word memory is absent. Small clinical samples, fixed stories, auditory channel selection, regional averaging and offline processing limit the test. The package preserves code, compact inputs, source versions and terms, all results, numerical checks, figures and repeat instructions. No shared positive memory effect or causal mechanism was established.

iEEG · fMRI · word repetition · auditory cortex · story listening · future control · negative result · independent confirmation