Psychometric Evaluation of the Artificial Intelligence Perception Scale for Cancer Patients: A Structural Equation Modeling Approach

Abstract

ABSTRACT Rationale, Aims and Objectives Artificial intelligence (AI) is increasingly integrated into oncology care, with oncology nurses playing a key role in patient education, communication, and trust in these technologies. Although artificial intelligence is increasingly integrated into healthcare, existing AI perception and technology acceptance scales have largely been developed for general healthcare settings and do not adequately capture the unique clinical, ethical, and workflow‐related aspects of oncology practice. This study aimed to develop and psychometrically validate a scale measuring oncology patients' perceptions of artificial intelligence. The study focused on instrument development rather than testing a theoretical model of technology acceptance. Methods A methodological, cross‐sectional scale development study. The scale was developed according to established nursing research guidelines. An initial item pool was generated using the Technology Acceptance Model, literature review and expert opinions from oncology nursing and interdisciplinary healthcare professionals. Content validity was assessed using the Davis technique and Lawshe method. Data were collected from 382 adult cancer patients receiving nursing care. Construct validity was examined using exploratory and confirmatory factor analyses. Reliability was evaluated using Cronbach's α, split‐half reliability and composite reliability. Structural equation modeling with bootstrapping tested the theoretical model. Results The final scale included 33 items across three factors: knowledge, awareness, and trust; perceived benefit; and acceptance and use tendencies. Exploratory factor analysis explained 82.37% of the total variance. Confirmatory factor analysis demonstrated good model fit (CFI = 0.99, RMSEA = 0.079, SRMR = 0.038). Internal consistency was excellent for the total scale (α = 0.97) and subscales (α = 0.97–0.98). Perceived benefit partially mediated the relationship between knowledge and acceptance, explaining 52% of the variance. Conclusions The scale is a valid and reliable instrument for assessing cancer patients' perceptions and acceptance of AI in oncology nursing practice. This scale provides oncology nurses and nurse researchers with a standardized tool to support nursing‐led patient education, shared decision‐making and the integration of AI into patient‐centred oncology nursing care.