The Academic Salon of the Department of Psychology and Cognitive Sciences is held every Thursday afternoon. Welcome to all students and faculty members from every department!
Time: June 12 (Thursday) afternoon 14:30
Location: Room 1110, 11th floor, Lv Dalong Building
Department of Psychology and Cognitive Sciences Xiao'ang Wan Research Group Academic Salon
Report One
Role Effects of Robot Service Providers: Consumers' Differentiated Responses to Automated Services in Restaurant Settings
Presenter: Xiyu Guo
Content:
With the development of artificial intelligence technology, the restaurant service industry is experiencing an automation revolution. More and more restaurants are beginning to introduce robot employees to optimize service processes and improve service quality. For example, robot waiters can reduce costs and enhance operational efficiency while creating novelty value. However, not all restaurant service positions are suitable for automation replacement. For instance, in core aspects involving food cooking, consumers still show significant insufficient acceptance of robot chefs. To deeply explore the impact of automation in different service positions on restaurant perceived value, we conducted two studies. In both studies, participants watched restaurant video advertisements featuring either robot or human waiters. Study one included human chefs, while study two included robot chefs. The results of both studies consistently showed that compared to human waiters, robot waiters could significantly enhance consumers' willingness to pay premium prices and strengthen restaurant innovation perception. Meanwhile, compared to human waiters, participants preferred dining at restaurants with robot waiters, but when robots served as chefs, the situation was reversed. Additionally, participants' acceptance of robot waiters was higher than that of human waiters. These results indicate that while front-of-house robot waiters may enhance perceived value through novelty, back-of-house robot roles may face consumer skepticism, highlighting the role-dependent nature of automation value in the hospitality industry. Our research results suggest that restaurants should use robots in customer-facing positions while retaining human employees in back-of-house operations to maximize business benefits.
Report Two
From Human Collective Wisdom to Artificial Intelligence Algorithms: How Advice Sources Influence Behavioral Health Risk Judgments
Presenter: Mengying Liu
Content:
With the widespread application of artificial intelligence (AI) in the health field, AI as an emerging information source for risk assessment and decision support, its reliability and influence are receiving increasing attention. However, it is currently unclear how individuals weigh and integrate advice from AI and human collectives under different threat levels of health contexts. This study systematically examined the differences between AI and human collectives in health risk assessment through two experiments, and explored how advice from these two sources influences individual risk judgments. In Experiment 1, 60 participants (gender balanced) and 30 AI samples conducted risk assessments of common health-related behaviors. Results showed that compared to human participants, AI exhibited significant risk overestimation tendencies, and this overestimation was mainly reflected in exaggerating outcome severity rather than risk probability. Experiment 2 adopted an advice-taking paradigm, where 60 participants, before and after receiving AI or human collective advice, made choices on paired high or low health threat behaviors (lower risk options) and reported choice confidence. The study found that in low-threat contexts, human collective advice showed stronger persuasiveness, with participants adopting advice from human collectives more than from AI; while in high-threat contexts, although there was no significant difference in adoption rates between the two types of advice, participants who accepted AI advice showed greater belief updating than those who accepted human collective advice. Overall, AI's risk assessment has systematic biases, but the negative impact of this assessment bias is reduced in high-threat contexts. This finding reveals how threat contexts influence human-AI information integration and provides insights for designing effective AI-based health decision support systems. Future AI system design should consider the impact of contextual factors on users' advice adoption and adopt differentiated advice presentation strategies for health problems of different threat levels.
