Factors influencing user experience in AI chat systems – a satisfaction study based on factor analysis and linear regression
JiaMan Xing,Qianling Jiang
DOI: https://doi.org/10.1108/k-10-2023-2237
IF: 2.352
2024-05-08
Kybernetes
Abstract:Purpose Since the introduction of the outstanding web AI chat system, ChatGPT, it has caused a significant impact in both academia and the business world. Many studies have started to explore its potential applications in various fields. However, there is a lack of research from the perspective of user experience. To fill this theoretical gap and provide a theoretical basis for the operation and design of related services, this study plans to develop a set of evaluation scales for AI chat system user experience and explore the relationship between various factors and user satisfaction. Design/methodology/approach This study obtained 41 evaluation indicators through literature review and user research. Subsequently, these indicators were used as questionnaire items, combined with satisfaction metrics. A total of 515 questionnaires were distributed, and factor analysis and linear regression were employed to determine the specific elements influencing user experience and the user satisfaction model. Findings This study found that the factors influencing user experience are usefulness, accuracy, logical inference, interactivity, growth, anthropomorphism, convenience, credibility, ease of use, creativity, and security. Among these factors, only accuracy, anthropomorphism, creativity, and security indirectly influence satisfaction through usefulness, while the rest of the factors have a direct positive impact on user satisfaction. Originality/value This study provides constructive suggestions for the design and operation of related services and serves as a reference for future theoretical research in this area.
computer science, cybernetics