Mean eye state log power gave no clear alpha agreement gain across one month with 280 selected seconds in 60 ds004148 adults

We tested whether an equal mean of eye state log alpha powers improves agreement across one month at fixed selected data time. We used 60 adults from OpenNeuro ds004148 version 1.0.1. The mean used 140 planned seconds per eye state. The comparator used 280 seconds with closed eyes. Absolute-agreement ICC was 0.887 for the mean and 0.831 for closed eyes. The difference was 0.056. Its 95% paired person-bootstrap interval was -0.015 to 0.157. The primary test did not establish a gain or equal agreement. Nine planned secondary checks showed effects of window choice, visit interval, reference, band, and scoring rule. The package preserves 360 spectra, 1,800 score rows, fixed scoring code, bootstrap outputs, source hashes, and reproduction instructions. Four raw recordings reproduced all saved arrays exactly. Two analysis runs gave identical result hashes. The data came from complete recordings with fixed state order and source processing. This analysis does not establish a shorter recording schedule, clinical value, or independent replication.

electroencephalography · alpha power · test-retest reliability · intraclass correlation · eye state · agreement · OpenNeuro