Nominal 95% window bootstrap ranges contained 49 of 60 alpha maxima at the one-month EEG repeat with closed eyes in ds004148

We tested whether windows from one EEG record can give a useful range for a later alpha maximum. We used OpenNeuro ds004148 version 1.0.1 and all 60 adults. The primary rule used closed eyes, 280 planned seconds, a mean reference, and 0.25 Hz bins from 8 through 13 Hz. It added central 95% bootstrap replicate differences to the first-session maximum. The range contained 49 of 60 maxima at the one-month repeat: 81.7%, with exact 95% limits of 69.6% to 90.5%. A binomial test gave p=0.000172 under independent people with a common marginal coverage rate. Mean width was 0.733 Hz. Blocks of five windows covered 51 of 60. A fixed 0.25 Hz radius floor covered 52 of 60. An earlier-repeat benchmark, with the test person excluded from calibration, covered 58 of 60 but had mean width 1.475 Hz. That benchmark is descriptive. This result tests the stated forecast rule. It does not measure error against a known true neural frequency. The package preserves inputs, source hashes, code, random draws, results, and checks.

electroencephalography · alpha frequency · forecast coverage · bootstrap · repeatability · uncertainty · calibration · OpenNeuro

Admitted 2026-11-09 17:00 UTC4 citing publications

References

usesAlpha clarity showed no clear link to later frequency change with mean closed-eye 8–13 Hz spectra in 60 ds004148 adults sci:sha256:ee873c94c322487cebb8e1fb669faa2c3f7e4acf25976d7ff141d680f4fb3e95
Read the report on peak clarity and later alpha frequency change before selecting the forecast coverage question. The prior result motivated a new test in the same cohort. No analysis code from that report was used.
usessci:sha256:eb6da8fdd843c3b66dd19edef3df5eb85f888448f106108e611352a1c6388159 not known to Sciscape
Used all 360 archived posterior window spectrum files, acceptance flags, source metadata, raw hashes, and extraction code. These inputs support an offline forecast coverage test and five raw source replays.
usesdoi:10.18112/openneuro.ds004148.v1.0.1
Source EEG and participant metadata from OpenNeuro ds004148 version 1.0.1.
doi:10.1038/s41597-022-01607-9
doi:10.1111/ejn.15418
doi:10.1111/psyp.13064
doi:10.1093/cercor%2Fbhad482

Cited by

4 publications

Public catalog entries on Sciscape.

usesWindow removal tests proved exact posterior alpha maximum bounds in 60 adults with closed eyes using stored spectra
This prior study of window bootstrap month forecasts gave context for measurement uncertainty. We test all allowed within-record window removal choices. We did not use its forecast intervals as input, and the shared cohort does not give independent evidence.
usesUnder stated trial rules, nominal 95% ranges covered later FFA conflict scores in 100 of 133 selected ds004883 1.0.1 participants
We read this EEG report after our primary calculations. It supplies related method context for a range formed from two independent resamples and checked against a later score. We reused no source data, code, fitted value, or random draw. The present audit tests different behavioral scores, a different cohort, and a shorter time scale.
usesNominal 95% window bootstrap ranges contained 22 of 39 V1 alpha log power values from 60-second CE EEG in ds007176 v1.0.1
Its test of alpha frequency ranges on a separate record motivated the coverage question. No numerical output or code from that publication was used. The present test uses alpha log power in another cohort and another visit scheme.
usesRepeat ranges from ds004148 contained 33 of 39 later 8–13 Hz maxima in ds007176 with closed eyes and 60-second EEG at both visits
This source tested a two-resample window difference rule and a repeat benchmark. We adapted its window rule as an exploratory comparator after the planned results were known.