#Nomask on Instagram: Exploring Visual Representations of the Antisocial Norm on Social Media

Abstract

Social media (SM) functions such as hashtags and photo uploading can enrich and expedite user interactions, but can also facilitate the online spread of antisocial norms. Mask aversion is one such antisocial norm shared on SM in the current COVID-19 pandemic circumstances. This study utilized the social representation theory (SRT) to explore how mask aversion is visually represented in the Instagram photos tagged with #NoMask. It examined the overall content of the photos, the characteristics of the faces portrayed in the photos, and the presented words in the photos. Additionally, the study grouped the photos through k-means clustering and compared the resulting clusters in terms of content, characteristics of the faces, presented words, pixel-level characteristics, and the public’s responses to the photos. The results indicate that people, especially human faces, were visually represented the most in the Instagram photos tagged with #NoMask. Two clusters were generated by k-means clustering—Text-centered and people-centered. The visual representations of the two clusters differed in terms of content characteristics and pixel-level attributes. The texts presented in the photos manifested a unique way of delivering key messages. The photos of the people-centered cluster received more positive comments than the text-centered one; however, the two clusters were not significantly different in eliciting engagement. This study can contribute to expanding the scope of SRT to visual representations and hashtag movements.