Key points are not available for this paper at this time.
Background This study aims to develop a valid and reliable scale for assessing individuals’ attitudes towards, perceptions of, and tendencies to engage in artificial intelligence-supported health counseling (AI-HC) and to examine its psychometric properties. Methods The research was conducted using a methodological design with a two-stage data collection process. The initial item pool consisted of 55 items, which was reduced to 32 items based on expert opinions. In the first stage, exploratory factor analysis (EFA) was performed with 592 participants, and in the second stage, confirmatory factor analysis (CFA) was conducted with 663 participants. Composite reliability (CR), average variance extracted (AVE), and the Fornell–Larcker criterion were calculated. Internal consistency was assessed using Cronbach’s Alpha and McDonald’s Omega coefficients. Data were analyzed using SPSS 26.0 and AMOS 24.0. Results In both the EFA and CFA datasets, the majority of participants were female and aged between 18–24 years. Most were university graduates, and “doctor” was the most frequently preferred source of health information. Participants’ knowledge levels regarding artificial intelligence technologies in health were generally reported as “low” or “moderate.” The socio-demographic distributions of both datasets were largely similar. EFA yielded a four-factor, 24-item structure (Usage and Trust, Privacy Perception, Supportiveness Perception, and Medical Competence). The Kaiser–Meyer–Olkin (KMO) value was 0.955, and the total variance explained was 66.01%. In CFA, fit indices (χ²/df = 2.957, goodness of fit index (GFI) = 0.918, adjusted goodness of fit index (AGFI) = 0.900, comparative fit index (CFI) = 0.945, normalised fit index (NFI) = 0.938, Tucker–Lewis index (TLI) = 0.919, root mean square error of approximation (RMSEA) = 0.054, root mean square residual (RMR) = 0.045) supported the model’s fit to the data. The four factors demonstrated convergent and discriminant validity, with CR and AVE values at acceptable levels, and inter-factor correlations supported discriminant validity. Cronbach’s Alpha and McDonald’s Omega coefficients ranged from 0.79 to 0.94 across all subscales, indicating high internal consistency. Based on the psychometric analyses, the scale was accepted as a valid and reliable 24-item, four-factor instrument. Conclusion The developed artificial intelligence-supported health counseling scale (AI-HCS) is a valid and reliable tool for measuring critical factors influencing the adoption of AI-HC. The scale can contribute to the development of AI integration strategies in health policies and the design of patient-centered digital solutions.
Ali GÖDE (Tue,) studied this question.