Abstract
Context and relevance. Studies of emotion perception based on facial expressions are primarily conducted using photo/video materials obtained from trained models or induced emotions under specially controlled conditions. It is clear that in everyday life situations, spontaneous emotions, expressed on the face through natural expressions, appear less expressive than staged prototypes. Consequently, in real-life communicative situations, emotion recognition is much more difficult, which can easily lead to erroneous conclusions by communication partners. The aim of this study was to investigate the nature of the perception of dynamic natural expressions demonstrated by participants in dyadic interactions in structured interview situations. Hypotheses were tested regarding the presence of specific assessment profiles: for each type of situation; for each model; for each video recording; for each observer. The main independent variables were: interview conditions and the natural expressions of the models. Dependent variables were the structure of observers' assessments. The stimulus material consisted of 35 videos of 7 sitters, one for each of the 5 interview conditions, each 5 seconds long. The primary assessment instrument was the Izard Differential Emotion Scale (DES — 33 Likert scales from 1 to 5). Results. Significant differences between conditions across all videos were obtained for 15 primary DES scales. The scales of interest, surprise, grief/sadness, disgust, contempt, shame, fear, and guilt were insignificant. Based on the obtained results, it can be assumed that the interview participants focused on the topics of the discussed issues that were unexpected for them, did not evoke approval, and put them in an awkward position. Cluster analysis distributed the entire set of assessments into 4 groups (k-means method). The variability between groups was at the level of 11% (SSB / SST = 0.111). Significant differences were found for virtually all DES scales, with the exception of individual scales related to the emotions of interest and contempt. Conclusions. It has been shown that natural expressions displayed by people in interviews are ambiguous and are perceived by observers inconsistently. However, despite individual variability in assessments, patterns are evident in the assessments made by naive observers, likely related to the context of the social situation, which sets the emotional tone of the communication partners.