ePoster

CHALLENGES IN GROUP-LEVEL STATISTICAL INFERENCE: COMPARING NONPARAMETRIC PERMUTATION AND PREVALENCE APPROACHES WITH EEG RESTING STATE

Jordi Tobajas-Arbósand 1 co-author

Universitat Internacional de Catalunya

FENS Forum 2026 (2026)
Barcelona, Spain
Board PS06-09PM-364

Presentation

Date TBA

Board: PS06-09PM-364

Poster preview

CHALLENGES IN GROUP-LEVEL STATISTICAL INFERENCE: COMPARING NONPARAMETRIC PERMUTATION AND PREVALENCE APPROACHES WITH EEG RESTING STATE poster preview

Event Information

Poster Board

PS06-09PM-364

Abstract

Group-level statistical inference in neuroimaging faces significant challenges driven by between-participant variability. Additionally, the dependent measures used to estimate individual averages often follow a non-normal distribution, such as heavy-tailed power spectra. These challenges are particularly pronounced in resting-state EEG, where cortical rhythms exhibit hallmarks of nonlinearity, such as alpha-band activity exhibiting high-amplitude “extreme” events. This poster implements previously established statistical pipelines written in MATLAB or Python by integrating multiple noisy withinparticipant replications into a single, stable global null test. Additionally, we quantify the proportion of the population exhibiting a true positive effect to provide a quantitative uncertainty estimate rather than a binary statistical inference. We apply and compare these frameworks by focusing on one of the most robustly replicated effects in human electrophysiology: the suppression of alpha power during eyes-closed compared to eyes-open conditions (alpha blocking). This approach establishes an important first step toward future applications in intracranial recordings, where robust statistical inference must be balanced against the limited and sparse electrode coverage typical of intracranial EEG experiments.

Recommended posters

Cookies

We use essential cookies to run the site. Analytics cookies are optional and help us improve World Wide. Learn more.