Measurement of General Attitude Toward AI and its Relationship with AI Anxiety, Attitude Toward Robots and Belief in Human Nature Uniqueness: Polish Adaptation of the ATTARI-12 Questionnaire

Abstract

Objective: This study introduces the ATTARI-12-PL, a Polish adaptation of the ATTARI-12 questionnaire designed to measure general attitudes toward artificial intelligence (AI) across cognitive, affective, and behavioral facets. Method: An online sample of 332 participants was used for the psychometric evaluation of the ATTARI-12-PL scale. The internal structure and consistency of the scale were assessed, and the validity of the questionnaire was evaluated through correlation analyses. Results: Single-factor model did not provide adequate fit in confirmatory factor analysis (CFA). In contrast, a bifactor model incorporating a general attitude factor alongside specific orthogonal factors (attitude facets and item wording) demonstrated very good fit indices and accounted well for the data structure. The scale showed high internal consistency (Cronbach's α = 0.94). Strong negative correlations were found between AI attitude and negative attitude toward interacting with robots, moderate negative correlations with AI anxiety and negative attitude toward robots with human traits, and a mild negative association with belief in human nature uniqueness (BHNU). The scale also showed a strong positive correlation with other AI measurement tool: AIAS-4-PL. Conclusions: The ATTARI- 12-PL is a reliable and valid tool for measuring general attitudes toward AI in the Polish language context. The scale's properties and significant correlations with related constructs support its use in future research on AI attitudes.