Posterior scalp alpha did not increase consistently with verbal memory load in 12 patients

This analysis tested whether posterior scalp alpha power increased with verbal working-memory load in epilepsy patients performing a Sternberg task. It used scalp EEG from 43 of 68 sessions in OpenNeuro ds004752 v1.0.1: all sessions of sub-01 through sub-10, four sessions of sub-11, and one session of sub-12. For correct, artifact-free trials with set sizes 4, 6, or 8, log10 mean 8–12 Hz power in the 3–6 s retention window was calculated with a Hann FFT, z-scored within session and channel, and averaged over posterior scalp channels after common-average referencing. Across 12 subjects, the mean Fisher z of the within-subject Spearman correlations was 0.069 (bootstrap 95% CI −0.045 to 0.182); 6 of 12 correlations were positive, and the exact two-sided sign-flip p was 0.27 (Holm-adjusted p = 0.54). The secondary Fz theta test included 7 subjects and was also not significant (mean z = 0.096, p = 0.38). The effect varied with reference and exploratory analysis choices. The planned SEEG analysis was not performed because the download stopped at a fixed time before results were known. These data do not show a consistent load-related increase in posterior scalp alpha in this sample, but the sample size and incomplete dataset do not establish absence of an effect.

working memory · scalp EEG · alpha power · memory load · Sternberg task · epilepsy · OpenNeuro ds004752 · sign-flip test