关键词: 生成式人工智能, 失真健康信息, 信息甄别, 用户体验, ChatGPT

Abstract: [Purpose/Significance] This paper aims to explore the role and user experience of using Generative Artificial Intelligence （GAI）in health misinformation identification from the perspective of technology-in-use. [Design/Methodology] This study used a mixed-methods approach incorporating fusion randomized controlled experiment, questionnaires, and interviews to gather research data, and analyzed the behavioral performance and perceived experience of users using GAI to identify health misinformation based on independent sample T-test and non-parametric test. [Findings/Conclusion] This research has found that the completion time for tasks in the GAI group are significantly shorter than that of the search engine group, while the average amount times of information interactions during the task completion process in the GAI group are reduced. The users' scores after using GAI search are significantly higher than their pre-search scores, and also significantly higher than those in the search engine group. The perceived effectiveness, efficiency, attractiveness, and overall experience of users in GAI group are significantly higher than those of the search engine group. There is no significant difference between the two groups in terms of perceived learnability, and error tolerance. The user experience of GAI in health misinformation identification follows the storyline of "task technology matching-user information behavior-user experience perception". [Originality/Value] This study comprehensively explore the application of GAI in health misinformation identification through a mixed-methods approach, which can provide insights for intelligent information services and interventions targeting health misinformation.

Keywords: Generative Artificial Intelligence, Health misinformation, Information identification, User experience, ChatGPT