Linear models gave an uncertain fixation theta difference after errors in 11 SEEG patients on trials with four letters

We tested hippocampal fixation theta power after an incorrect previous response. We used public SEEG data from OpenNeuro ds004752, version 1.0.1. The primary sample had 11 patients, two sessions per patient, and 377 current trials with four letters. Of these trials, 74 followed an error. Patient linear models adjusted for session, trial order, and previous load. The equal-patient mean difference was -1.05 dB, with a 95% t interval from -2.48 to +0.39 dB. The exact two-sided sign-flip test gave p = 0.1357. This test requires independent patients and symmetric null patient effects. It did not establish a group difference or prove that the difference was zero. Errors were observed outcomes, so the estimate does not establish a causal effect. The package preserves code, spectra, raw examples, source identities, results, and reproduction instructions. Fresh retrieval checked all 1,167 fixation windows against pinned public files. Independent calculations checked features, patient estimates, and the primary test.

hippocampus · theta power · stereoelectroencephalography · working memory · error history · fixation baseline · sign-flip test