Predicting Big Five Personality Traits from Resting-State EEG Connectivity: a Connectome Predictive Modeling Study

Personality traits are stable individual differences linked to important life outcomes including mental health, occupational functioning, and interpersonal relationships, yet their neural bases remain poorly characterized. Prior electroencephalography (EEG) studies have mainly examined isolated features such as spectral power or frontal asymmetry and have yielded mixed results. Building on recent work in connectome based predictive modeling, this study tested whether resting state EEG connectivity can predict Big Five personality traits in a nonclinical student sample. Resting state EEG (eyes open and eyes closed; about five minutes) was recorded from 115 healthy Chinese university students aged 18 to 28 years, who completed a psychometrically validated Chinese version of the Big Five Inventory 2. The exploratory CPM analyses reported here were conducted using eyes-open connectivity matrices derived from amplitude envelope correlation (AEC) and debiased weighted phase lag index (dwPLI) across canonical frequency bands. At the model-specific level, three CPM models yielded modest cross-validated correlations between observed and predicted trait scores: alpha-band AEC for Conscientiousness (r = 0.30, 95