Wong, NormanLookadoo, Kathryn2017-07-252017-07-252017-08-01http://hdl.handle.net/11244/51839The risk convergence model (RCM) was developed as a theoretical framework to organize research on media effects and narrative persuasion (So & Nabi, 2013). The model identifies social distance to a fictional character as a key mechanism through which narrative engagement variables impact media users’ personal risk perceptions. However, the model has only been tested twice (So & Nabi, 2013; So & Shen, 2015). This dissertation expands the RCM by adding parasocial relationship (PSR) with the character to the model. Additionally, the study tests the effects of character valence and affective disposition on social distance with an at-risk character. This dissertation also explores the influence of narrative ending types on risk perceptions. Finally, the study examines the duration of effects for personal risk perceptions. An experiment was conducted in which participants (N = 272) were randomly assigned to a character valence condition (positive or negative) and a narrative ending condition (uncertain, denial, or apology). Results indicated that a reduction in social distance to an at-risk character led to risk convergence. The study supported the addition of PSR with the character to the RCM because social distance mediated the relationship between PSR and character- and self-risk discrepancy. Character valence did not work with the narrative engagement variables, whereas results for affective disposition indicated a future role it could play in the RCM. Finally, narrative ending types did not significantly affect risk perceptions. Overall, this study’s results offer support for the RCM and offer new components to consider when testing the model in the future.Communicationrisk communicationnarrative engagement