Emotion-Antecedent Appraisal Checks: EEG and EMG data sets for Novelty and Pleasant- ness Jacobien M. van Peer 1 , Eduardo Coutinho 2,3 , Didier Grandjean 4 and Klaus R. Scherer 4 1 Behavioural Science Institute, Radboud University Nijmegen, Nijmegen, Netherlands 2 Department of Music, University of Liverpool, Liverpool, United Kingdom 3 Department of Computing, Imperial College London, London, United Kingdom 4 Swiss Center for Affective Sciences, University of Geneva, Geneva, Switzerland December 2, 2017 T his document describes the full details of the first data set (Study 1) used in Coutinho et al., to appear. The Electroencephalogra- phy (EEG) and facial Electromyography (EMG) sig- nals included in this dataset, and now made public, were collected in the context of a previous study by Peer, Grandjean, and Scherer, 2014 that addressed three fundamental questions regarding the mech- anisms underlying the appraisal process: Whether appraisal criteria are processed (a) in a fixed se- quence, (b) independent of each other, and (c) by different neural structures or circuits. In that study, an oddball paradigm with affective pictures was used to experimentally manipulate novelty and in- trinsic pleasantness appraisals. EEG was recorded during task performance, together with facial EMG, to measure, respectively, cognitive processing and efferent responses stemming from the appraisal manipulations. In comparison to the data collected and analysed in the original study (Peer, Grandjean, and Scherer, 2014), this dataset contains some differences in both EEG and EMG signals. This is due to changes in the pre-processing steps (i.e., the processing of the raw data), which have had an impact on the signals them- selves and also led to the removal of some trials. Full details, including information about data collection, are provided in the following subsections. Participants Twenty-six right-handed healthy students from the Uni- versity of Geneva (12 men, 14 women) participated for financial compensation (45 Swiss francs (CHF)). Inclu- sion criteria were age 18–35 years, right-handedness, excellent understanding of French, normal vision (no glasses or contact lenses), and good general health (no use of medication, except oral contraceptives). Ex- clusion criteria were psychological problems, a history of neurological disorders or head trauma, and use of hard or soft drugs. All participants provided written in- formed consent prior to their participation in the study, which was approved by the local ethical committee. Materials Stimuli for the oddball task (Fig. 1) consisted of 78 pleasant, 78 unpleasant, and 78 neutral color pictures selected from the International Affective Picture System database Lang, Bradley, and Cuthbert, 2001. To avoid low-level visual effects on EEG measures, all pictures were automatically corrected to an average luminance value of 0.48, using MATLAB software (version 7.10.0). The corrected pictures were visually inspected to make sure they did not look strange. Statistical analysis further confirmed that the three categories did not differ in high or low spatial frequencies (using the method described in (Delplanque et al., 2007), all F (2, 222) < 1.37,p>.25, or in number of humans and faces (human: Chisq =2.83,p =0.24; face: Chisq = 2.27,p =0.32). Statistical analysis of the normative subjective ratings of the pictures showed that the three categories differed significantly in valence (all pairwise comparisons, F (1, 148) > 1282.0,p<.001), and that neutral pictures were significantly less arousing than pleasant and unpleasant pictures (both F (1, 148) > 153,p<.001), but pleasant and unpleasant pictures did not differ significantly on arousal (F (1, 148) = 0.70,p = .40).