The two trial hippocampal theta error test did not reject zero after correct intervening responses at load four in 13 SEEG patients under patient independence and null sign symmetry

We tested hippocampal theta two trials after errors in public SEEG data. Each history had a correct intervening response at load four. OpenNeuro ds004752 version 1.0.1 supplied 610 histories from 13 patients and 57 sessions. The mean error contrast was -0.099 dB. The 95 percent patient bootstrap interval was -0.613 to +0.567 dB. The exact two sided test gave p = 0.800537. It did not reject zero at 0.05 under patient independence and null sign symmetry. This result does not establish no effect. A matched descriptive check gave -0.968 dB at the intervening fixation. The paired later minus intervening contrast was +0.869 dB, with an interval from +0.131 to +1.622 dB. This pattern cannot establish causal recovery. Errors were observed, and selection required a correct intervening response. We reused existing features and supplied a new trial history analysis. The package retains code, features, event records, complete fits, uncertainty draws, source versions, and retrieval instructions. Separate code checked all 104 patient models. A complete repeat matched 114 result file hashes. New raw checks passed for 456 windows across all used sessions and three native sample rates. A full raw session download reproduced every feature exactly. The preserved plan has a dated hash. The report corrects its assumption that every error forced the next load to four. The correction did not change the fixed sample or test.

hippocampal theta · SEEG · verbal working memory · trial history · error processing · spectral analysis · sign flip test