Simmonds, Benjamin2024-07-092024-07-092022https://hdl.handle.net/2440/141572This item is only available electronically.With the rise of social media has come a rise in the accessibility of health-related information from a variety of sources, which may vary in accuracy and agreement. To navigate conflicting information, people are often required to rely on readily available cues to consensus. However, it remains unclear which consensus cues can make information persuasive. It is also unclear whether an expert minority is sufficient to outweigh a conflicting majority view that is presented by laypeople. To explore these questions, an adult sample (N = 101) drawn from Amazon's Mechanical Turk was exposed to eight health claims within a mock Twitter platform. Participants made belief ratings for each claim before and after exposure to a series of Pro and Con response tweets to the claim, measuring belief revision. Participants were also prompted to choose one response to "retweet", measuring sharing behaviour. The expertise of some of the tweet authors varied within subjects but between trials. Within trials, certain tweet authors and arguments were frequently repeated. Results showed that in the absence of expert authors, participants favoured stances supported by a larger number of tweets. When experts were present, however, experts were favoured even when outnumbered in tweet quantity. These findings were consistent across both measures of belief revision and sharing behaviour. However, no significant effect on sharing behaviour was found when considering author or argument frequency. These findings are important in understanding how socially acquired information (and misinformation) shifts opinion and is spread, and the role that experts can play.Honours; PsychologyThesis