How do Individuals With and Without Traumatic Brain Injury Interpret Emoji? Similarities and Differences in Perceived Valence, Arousal, and Emotion Representation

Abstract

Impaired facial affect recognition is common after traumatic brain injury (TBI) and linked to poor social outcomes. We explored whether perception of emotions depicted by emoji is also impaired after TBI. Fifty participants with TBI and 50 non-injured peers generated free-text labels to describe emotions depicted by emoji and rated their levels of valence and arousal on nine-point rating scales. We compared how the two groups’ valence and arousal ratings were clustered and examined agreement in the words participants used to describe emoji. Hierarchical clustering of affect ratings produced four emoji clusters in the non-injured group and three emoji clusters in the TBI group. Whereas the non-injured group had a strongly positive and a moderately positive cluster, the TBI group had a single positive valence cluster, undifferentiated by arousal. Despite differences in cluster numbers, hierarchical structures of the two groups’ emoji ratings were significantly correlated. Most emoji had high agreement in the words participants with and without TBI used to describe them. Participants with TBI perceived emoji similarly to non-injured peers, used similar words to describe emoji, and rated emoji similarly on the valence dimension. Individuals with TBI showed small differences in perceived arousal for a minority of emoji. Overall, results suggest that basic recognition processes do not explain challenges in computer-mediated communication reported by adults with TBI. Examining perception of emoji in context by people with TBI is an essential next step for advancing our understanding of functional communication in computer-mediated contexts after brain injury.

Classification of Fashion Models’ Walking Styles Using Publicly Available Data, Pose Detection Technology, and Multivariate Analysis: From Past to Current Trendy Walking Styles

“…We therefore set the minimum number of data units to be included in each cluster at 7, which is approximately 10% of the total number of data units analyzed (i.e., 69 in total). As previous studies have and several lectures have instructed, we determined the optimal number of clusters comprehensively from the percent changes of the agglomeration schedule coefficients at each stage, in combination with the visual inspection of the dendrogram [ 37 , 38 , 39 ]. A hierarchical cluster analysis produced the dendrogram shown in Figure 1 and the agglomeration schedule coefficients presented in Table A2 ( Appendix C ) as outputs.…”