The Illusion of Knowing: Dunning-Kruger Effect in Generative AI Misinformation Detection
The Illusion of Knowing: Dunning-Kruger Effect in Generative AI Misinformation Detection
Mengqi (Maggie) Liao, Chuan, C., Li, W., & Sun, R. (2026, August). “The Illusion of Knowing: Dunning-Kruger Effect in Generative AI Misinformation Detection,” paper to be presented at the 109th annual conference of the Association for Education in Journalism and Mass Communication (AEJMC), New Orleans, LA.
Abstract: This study examines the Dunning-Kruger effect in the context of generative AI literacy, focusing on how users assess and evaluate the accuracy of AI-generated health information. Using a two-wave pre-registered design (N=506), results reveal a systematic mismatch between perceived and actual performance: low performers overestimate their abilities, whereas high performers underestimate them. This relationship is further moderated by users’ domain knowledge about the health topic (i.e., cancer knowledge). Implications of the findings are discussed.
Related Research
-
Interventions that Matter: Researcher-Practitioner Partnership in Intimate Partner Prevention in Immigrant CommunitiesJeong-Yeob Han, “Interventions that Matter: Researcher-Practitioner Partnership in Intimate Partner Prevention in Immigrant Communities,” presentation to U.S. Department of Justice Office on Violence Against Women (OVW) Learning Hour, May 20, […]
-
Ready for the Dark Side of the Moon: Towards a Human-AI-collaborated ‘Triple-Check’ Crisis Information Governance ModelYan Jin (invited speaker). “Ready for the Dark Side of the Moon: Towards a Human-AI-collaborated ‘Triple-Check’ Crisis Information Governance Model.” Rotterdam School of Management Seminar: “AI and Crisis Management: Human […]