The theta score between adjacent trials was 0.004 in 12 SEEG patients after task, order, and Gaussian bias correction

We tested hippocampal theta power between adjacent source trials in public OpenNeuro ds004752 version 1.0.1. The primary sample had 12 patients, 24 sessions, 1,000 correct trials without artifact flags, and 846 adjacent links. We adjusted for task conditions and quadratic trial order. Independent Gaussian errors passed through the same regression and mask supplied a bias correction. The equal patient mean corrected score was 0.0042. Its descriptive 95 percent t interval was -0.0609 to 0.0693. The exact two-sided patient sign flip probability was 0.8857. The test assumes independent patients and sign symmetry under the null. Raw residual correlation was -0.0708; the Gaussian null mean was -0.0750. Five fixed checks and later sessions did not show a clear mean relation. Later sessions reused nine patients. These results do not establish trial independence. The package preserves code, spectra, source versions, raw examples, simulations, independent checks, figures, and reproduction instructions.

stereoelectroencephalography · hippocampus · theta power · serial correlation · regression bias · verbal working memory