AI Recommendation Service Acceptance: Assessing the Effects of Perceived Empathy and Need for Cognition

Autor: Namhee Yoon, Ha-Kyung Lee
Jazyk: angličtina
Rok vydání: 2021
Předmět:
Zdroj: Journal of Theoretical and Applied Electronic Commerce Research, Vol 16, Iss 5, Pp 1912-1928 (2021)
Druh dokumentu: article
ISSN: 0718-1876
DOI: 10.3390/jtaer16050107
Popis: This study investigated the effect of perceived technology quality and personalization quality on behavioral intentions, mediated by perceived empathy in using an artificial intelligence (AI) recommendation service. The study was based on a theoretical model of artificial intelligent device use acceptance. We also tested the moderating effect of individuals’ need for cognition, influencing empathy. Data collection was conducted through an online survey using a nationally recognized consumer research panel service in Korea. The participants were asked to respond to their preferences and needs on sneakers; then, they randomly experienced the AI (versus human expert) recommendation service that offers a recommended product. A total of 200 data were analyzed using SPSS 21.0 for descriptive statistics, reliability analysis, and PROCESS analysis, and AMOS 21.0 for confirmatory factor analysis and structural equation modeling (SEM). Results revealed that, compared with the human (expert) recommendation service, the AI recommendation service increased perceived technology quality, which increased personalization quality. Technology and personalization quality had a positive influence on behavioral intentions, mediated by perceived empathy. In addition, when individuals had a high level of need for cognition, the effect of personalization quality on empathy was stronger. However, individuals with a low level of need for cognition perceived greater empathy, as technology quality increased. The findings of the current study improve understanding of how consumers accept AI technology-driven services in the online shopping context.
Databáze: Directory of Open Access Journals