The Devil's in the Variability: A Multidimensional Analysis of EEG Signal Dynamics and Their Relation to Behaviour

Electroencephalographic (EEG) recordings offer high temporal resolution, a key advantage for understanding the variable and time-evolving nature of brain activity. In recent years, interest in variability and complexity measures has grown, as traditional EEG methods might not capture the full informational value of EEG recordings. Yet, it remains unclear whether different variability measures offer reliable and valid methods for quantifying fluctuations in brain activity. To address these issues, we computed different amplitude- and frequency-based variability measures for EEG data recorded during the performance of two task-switching paradigms across two independent measurement sessions. Results indicated that variability measures shared a significant amount of variance and were highly reliable both within and across measurement sessions. In addition, on/off-task effects (i.e., changes from baseline to overt task performance) were considerably larger than task-switching-specific effects.