Psychometric evaluation of the Turkish version of the AI Chatbot Dependence Scale (AICD-8)

The rapid integration of generative artificial intelligence-based chatbots into daily life has intensified discussions about their potential association with problematic use and addiction-like behaviors. In this context, several psychometric scales have been developed to assess problematic chatbot use including the eight-item AI Chatbot Dependence Scale (AICD-8). However, the AICD-8 has not been validated into Turkish. Therefore, the present study translated the AICD-8 into Turkish, and its psychometric properties were evaluated. The sample comprised 600 individuals (249 employees and 351 university students). Descriptive statistics, reliability indicators (Cronbach’s alpha, McDonald’s omega, and composite reliability), Pearson correlation coefficients, and confirmatory factor analysis (CFA) were conducted to examine the psychometric properties of the scale. The Turkish version of the scale (AICD-8-T) demonstrated good internal consistency, and CFA confirmed that the single-factor model exhibited acceptable fit. Convergent validity was supported through strong associations between the AICD-8-T and the Problematic ChatGPT Use Scale. Nomological validity was evidenced by expected correlations with social media addiction, loneliness, and internet addiction. Criterion-related validity was supported by strong positive correlations with daily ChatGPT use and chatbot use. Overall, the findings indicate that the AICD-8-T possesses robust psychometric qualities and can be reliably used in future psychological research in Türkiye.