Validation of the AI mental health Therapy Attitude, Perception, and Intention Scale (AiTAPI).

BACKGROUND

This study aims to validate the newly designed Attitude Perception and Intention Scale for adopting AI Mental Health Therapy by employing the Classical Test Theory (CTT) approach.

METHODS

Items formation followed standard procedures with the calculation of the Content Validity Index. The testing and validation samples were gathered through online and face-to-face cross-sectional survey using a self-reported questionnaire. The scale underwent Exploratory Factor Analysis (EFA) and Confirmatory Factor Analysis (CFA). Other psychometric examinations were applied for the reliability and convergency/divergency validity of the three subscales.

RESULTS

Factor analyses confirmed that a three-factor model of the scale fit the data well, with the highest Eigenvalue of 4.43 for the attitude subscale, explaining 84.4% of the total variance. Model fit statistics indicated the three-factor model fitted well with the data. Both attitude and intention subscale demonstrated satisfactory convergent and discriminant validity in the post CFA analysis, but not for the perception subscale. Both the attitude and intention subscales demonstrated high internal consistency, with Cronbach's alpha values of 0.920, and 0.828 respectively with the inter-item correlation of 0.647 for the perception subscale. All three subscales correlated significantly with mental well-being and coping measures.

CONCLUSIONS

The results indicated that a three-factor model fits best to the current data. Two of the three subscales have good reliability and validity for measuring the attitude and intention of using AI Mental Health Therapy. However, the perception subscale requires further development.