I助手用户信任的关键影响因素及其效应的元分析

Purpose/Significance] With the rapid development of artificial intelligence (AI) technology, the penetration rate of AI assistants continues to increase in industries such as retail, finance, and healthcare. The influence mechanism of users on their usage intention has also attracted extensive attention from scholars. User trust is an important driving factor for the adoption of the product, but the existing research has not yet reached a consensus on the key influencing factors of trust and their effects, and a systematic review and evaluation of these factors is urgently needed. [Method/Process] The study takes 65 empirical research papers published at home and abroad as samples. We have quantitatively analyzed the influencing factors of user trust for AI assistants and its relationship with usage intention by using meta-analysis. Then, we test the moderating effect of user cultural background and product usage scenarios on this relationship. [Result/Conclusion] The results show that product anthropomorphic traits, perceived usefulness, perceived ease of use, social presence, social influence, product performance and perceived risk are seven important factors that affect user trust. Among them, except perceived risk has a significantly negative impact on user trust, the other factors have a significantly positive impact on user trust. In addition, user trust has a significantly positive impact on usage intention, but there is some heterogeneity among different studies. Through moderating effect analysis, it is found that user cultural background and AI assistant product use scenarios are the main reasons for heterogeneity. The research findings provide a theoretical basis for future in-depth studies and offer guiding suggestions for the interactive design and application promotion of AI assistant products.