Explainable Medical AI: The Impact of Agent and Explanation Types on Health Persuasion

Abstract

Despite the rapid advancements in medical artificial intelligence (AI), little is known about how different types of explanations influence user compliance in healthcare settings. This research examines the impact of agent types (AI vs. human) and explanation types (mechanistic vs. teleological) on health persuasion. Through three studies conducted in diverse health contexts (HPV vaccination, breast cancer screening, and herpes zoster vaccination), we found that AI agents using mechanistic explanations significantly enhance users' self-efficacy, leading to higher health compliance, while human agents employing teleological explanations improve users' response efficacy, resulting in increased health compliance. This research contributes theoretically to the literature on explainable AI, AI agents, and health communication and has practical implications for healthcare marketers by demonstrating how matching explanation types with agent types can optimize health persuasion.

Ähnliche Arbeiten