Zhang Xing, Yin Mingyue, Zhang Mingyang, Li Zhaoqian, Li Hansen
Department of Physical Education and Sport, Faculty of Sport Sciences, University of Granada, Granada, Spain.
School of Athletic performance, Shanghai University of Sport, Shanghai, China.
Cyberpsychol Behav Soc Netw. 2025 Feb;28(2):126-131. doi: 10.1089/cyber.2024.0240. Epub 2024 Nov 26.
In recent years, a plethora of artificial intelligence (AI) chatbots have been developed and made available to the public. Consequently, an increasing number of individuals are integrating AI chatbots into their daily lives for various purposes. This trend has also raised concerns regarding AI chatbot dependence. However, a valid and reliable scale to assess AI chatbot dependence is yet to be developed. Therefore, this study was designed to develop and validate an AI chatbot dependence scale. We obtained initial items from previous publications and in-depth interviews. Subsequently, item analysis, exploratory factor analysis (EFA), confirmatory factor analysis (CFA), reliability, and validity analyses were performed to validate the AI chatbot dependence scale. Seventeen items underwent item analysis and EFA, resulting in a single-factor model with eight items explaining 58.42% of the total variance. The CFA indicated that our AI chatbot dependence scale had acceptable model fitting indices, with standardized loadings ranging between 0.50 and 0.76. In addition, this scale exhibited good reliability and validity. Thus, the current AI chatbot dependence scale can effectively evaluate individuals' dependence on AI chatbots in their daily lives.
近年来，大量的人工智能（AI）聊天机器人已被开发并向公众开放。因此，越来越多的人出于各种目的将AI聊天机器人融入他们的日常生活。这一趋势也引发了对AI聊天机器人依赖的担忧。然而，一个有效且可靠的评估AI聊天机器人依赖程度的量表尚未开发出来。因此，本研究旨在开发并验证一个AI聊天机器人依赖量表。我们从先前的出版物和深入访谈中获取了初始项目。随后，进行了项目分析、探索性因素分析（EFA）、验证性因素分析（CFA）、信度和效度分析，以验证AI聊天机器人依赖量表。17个项目进行了项目分析和EFA，得到了一个单因素模型，其中8个项目解释了总方差的58.42%。CFA表明，我们的AI聊天机器人依赖量表具有可接受的模型拟合指数，标准化载荷在0.50至0.76之间。此外，该量表表现出良好的信度和效度。因此，当前的AI聊天机器人依赖量表可以有效地评估个体在日常生活中对AI聊天机器人的依赖程度。
