An exact test did not detect increased hippocampus–middle temporal gyrus theta phase association during late memory delay

We tested phase association between hippocampus and middle temporal gyrus in public verbal memory recordings. The source was OpenNeuro ds004752 version 1.0.1. We used 11 SEEG patients, two sessions per patient, and 1,001 trials without marked artifacts. Both regions had eligible contact pairs on the same electrode shafts. We compared fixation with the last second of memory delay. The phase measure used Fourier bins at 4, 5, 6, and 7 Hz. Each score subtracted the exact mean over arbitrary trial pairing within each memory load. The mean late minus fixation change was -0.00248 excess pairwise phase consistency. The 95% Student t interval was -0.00927 to +0.00432. The exact two-sided patient sign test gave p = 0.453125. The test assumed independent patients and symmetric null signs. It did not detect an increase. It did not establish equivalence or absence of association. All 2,140 source windows were freshly checked against fixed public records. All digital hashes and signal-limit flags matched. The maximum relative coefficient error was 0. A fresh process reproduced all 13 analysis files byte for byte. The package preserves code, compact data, provenance, proof, results, and repeat instructions. The trial reference method was reused from earlier bilateral hippocampal work. This work adds a distinct anatomical comparison. The local plan preceded the new phase scores. It was not an independent registration.

theta phase association · pairwise phase consistency · hippocampus · middle temporal gyrus · verbal working memory · SEEG · exact sign test