Disentangling valence and arousal effects during emotion word processing

Abstract

Emotional content affects early processing of words. Scott et al. (2009) manipulated word valence and frequency in a lexical decision task (LDT) in an ERP study. They found P1 was greater for frequent positive and neutral words than frequent negative words and a reversed interaction at N1 with an early Posterior Negativity that was greater for valenced than neutral words. However, emotional arousal of stimuli was not manipulated. The aim of this study was to disentangle the effects of both valence and arousal. 28 English native speakers performed a LDT while EEG was recorded. 150 words (Citron et al., in press) were presented: 50 positive, 50 negative, and 50 neutral. Half of the valenced words were high (HA) and half low (LA) in arousal. EEG was recorded using a 128-channel Net (EGI, Eugene, Oregon). Impedance was kept below 50 k?. Sampling rate was 250 Hz. EEG was band-pass filtered between 0.3 and 40 Hz and segmented from 100 ms before to 1300 ms after stimulus onset. Segments were baseline corrected and re-referenced to the mastoids. Reaction times (RTs) showed main effects of valence and arousal and an interaction between arousal and valence.