ABSTRACT
This study analyzed Twitch chat messages for verbal indicators of the presence of parasocial relationships. Unlike traditional parasocial research, this study employed natural language processing to score streamer-targeted and viewer-targeted messages for verbal immediacy. It divided chat data according to stream content and streamer type and found that streamer-targeted messages consistently scored higher in verbal immediacy than viewer-targeted messages. The verbal immediacy scores for this dataset were content-agnostic. The findings illustrated a new method for testing the perceived relational closeness of parasocial relationships, namely, utilizing user-generated content to identify verbal indicators of parasocial relationships. Researchers are now capable of exploring the variance of parasocial relationships as they are naturally presented through new media platforms, where media users and figures co-exist.
Disclosure statement
No potential conflict of interest was reported by the author(s).
Note on contributor
Alex P. Leith is an assistant professor in the Department of Mass Communications at Southern Illinois University Edwardsville. Their research involves the application of computational analytics to online community and video game research. Areas are particular interest are relationships, mental health, and toxicity.