Blame the messenger: perceived mis/disinformation exposure on social media and perceptions of newsfeed algorithmic bias. Issue 2 (4th May 2021)

• Record Type: Journal Article
• Authors: Jardine, Eric
• Is Part Of: Journal of cyber policy. Volume 6:Issue 2(2021)
• Page Start: 196
• Page End: 213
• Publication Date: 2021-05-04
• Subjects: Misinformation -- algorithmic bias -- social media -- Facebook -- Twitter -- fake news; Cyberspace -- Periodicals; Internet -- Periodicals; Cyber intelligence (Computer security) -- Periodicals; 004.6
• DOI: 10.1080/23738871.2021.1953557
• Languages: English
• ISSNs: 2373-8871

Abstract

ABSTRACT: Does perceived exposure on social media to mis/disinformation affect user perceptions of social media newsfeed algorithmic bias? Using survey data from eight liberal democratic countries and propensity score matching statistical techniques, this paper details the average treatment effect (ATE) of self-reported perceived exposure to mis/disinformation on perceptions that social media newsfeed algorithms are biased. Overall, the results show that self-reported perceived exposure to misleading content on social media increases perceptions of algorithmic bias. The results also detail interesting platform/country variation in the estimated average treatment effect. The ATE of perceived fake news exposure on perceptions of algorithmic bias are similar on Twitter and Facebook but are amplified in countries with high society-wide issue salience surrounding 'fake news' and, especially, 'algorithmic bias'.