Impact of algorithm-driven comments on corrective information among young adults: focusing on emotional tone of comments and misinformation credibility

Purpose When facilitating transmission of health information from government officials to the public, social media employs algorithms that selectively expose users to specific perspectives, even for accurate health-related information from official sources. The purpose of this study was to explore impact of algorithm-driven comments characterized by different emotional tones (i.e. positive vs. negative vs. mixed) on users’ perceptions of credibility of corrective information to examine misinformation about flu vaccines aimed at young adults. Additionally, this study explored how prior misinformation credibility acted as an intervening variable in shaping the impact of algorithmically generated comments with diverse emotional tones on credibility of corrective information, with algorithm credibility serving as a mediator. Design/methodology/approach An online experiment was conducted with 275 participants recruited from Amazon Mechanical Turk (MTurk). Young adults from the USA aged between 18 and 35 years who were also users of Instagram were eligible for participating in this study as this study utilized Instagram platform for stimuli. Findings Results highlighted a diminished impact of algorithm-generated negative comments on perceived credibility of corrective information. Additionally, individuals with high misinformation credibility demonstrated a stronger tendency to trust algorithms featuring negative comments, underscoring the significant impact of algorithm-driven negativity in shaping trust dynamics for this group. Notably, credibility of the algorithm among individuals with high misinformation credibility did not translate into increased credibility for corrective information. This suggests that strategically designing algorithms to emphasize supportive or diverse opinions can be an effective approach to alleviate potential negative consequences associated with accurate information. Originality/value This research signifies the initial effort to disentangle the dynamics between negativity bias and cue routes within the algorithmic framework, shaping individuals’ perceptions of credibility of accurate health-related information contingent on accompanying comments. In the context of social media platforms that embrace diverse opinions, it emphasizes the critical necessity for tailored algorithmic strategies to effectively deliver accurate information.