The overlap-emotional task

The overlap-emotional task was performed to measure the accuracy and reaction time for categorizing the target emotion in context (see Figure 1). Contextual emotions might be the same or different from the target emotion (Righart and de Gelder, 2008). All pictures of the faces were used in full-blown facial expressions. The face stimuli were composed of 24 color face photographs balanced for gender, taken from “Extended Chae Lee Korean Facial Expressions of Emotions: ChaeLee-E” (Lee et al., 2013). The facial expressions included fearful (8), sad (8), and happy (8). The pictures of context consisted of 24 pictures; fearful (8), sad (8), and happy (8) scenes taken from the International Affective Picture System (Lang et al., 1997), the Open Affective Standardized Image Set (Kurdi et al., 2017), and the Geneva Affective Picture Database (Dan-Glauser and Scherer, 2011). Supplementary Table 1 shows which pictures are used. To validate the valence and arousal of the pictures displaying fearful, sad, and angry expressions, 15 graduate students rated the relevance, arousal, and valence of the emotions using a seven-point Likert scale (0 = Not at all, 6 = Extremely). Notably, there was no significant difference in the arousal of emotions for either type of stimulus. Supplementary Table 2 shows the ratings for fearful, sad, and angry emotions of the pictures.

Trial example of the overlap-emotional Task. The stimuli were presented under the condition of congruence or incongruence. The original stimuli are unable to be reproduced for copyright reasons but are available upon request from the corresponding author.

The participants were required to focus on the center of the target-context stimulus when they performed the task. The facial stimuli were placed at the center of the scene to avoid saccades. Emotional scenes were carefully edited to prevent the facial stimuli from covering a critical part of the scene related to emotion. The width and height of the target (emotional faces) and context (emotional scenes) were 7 × 9, and 35 cm × 26 cm, respectively. After fixation was presented for 1 s, the stimuli were presented only for 0.2 s and disappeared. The participants were required to answer what the emotion of the target was as quickly and accurately as possible while ignoring the emotions of the surrounding context. When the targets were presented as fearful, sad, or happy, the answers were the F, S, or H, buttons, respectively. They were instructed to react with their hand on the keyboard to respond as quickly as possible. If a participant failed to respond within 2.5 s, it would be an incorrect answer and moved on to the next question. Reaction time was rated from the point at which the stimulus was presented. This task consisted of eight blocks, with each block composed of 48 trials. The total number of congruence and incongruence conditions was the same to avoid predictability (FF, SS, and HH, each of eight trials per block; FS, FH, SF, SH, HF, and HS, each of four trials per one block), and the order of each trial was presented randomly.