Development and validation of the scale for attitudes towards generative AI (SAGAI)

Abstract

The Scale for Attitudes towards Generative Artificial Intelligence (SAGAI) was developed to understand learners' attitudes and perceptions towards the use of generative AI technologies in educational settings. Grounded in theoretical frameworks such as technology acceptance, planned behaviour, diffusion of innovations and social identity, the scale focuses on capturing attitudinal dimensions—including perceived usefulness, expectancy, competency and anxiety—rather than directly measuring behavioural engagement. The instrument was created through a systematic process beginning with an extensive item pool informed by literature and theory, followed by expert review and pilot testing. Its validity and reliability were examined through exploratory factor analysis with 244 undergraduate students and subsequently cross-validated via confirmatory factor analysis with another sample of 243 students. The analyses resulted in a 23-item scale comprising four distinct factors, each reflecting a different aspect of learners' attitudes towards interacting with generative AI. Findings indicated that students generally held positive perceptions about GenAI's benefits and future potential, although some degree of apprehension persisted, particularly reflected in higher anxiety scores. Overall, SAGAI offers a reliable and valid tool for gaining insights into learners' attitudes, competencies and concerns regarding GenAI's role and integration in education, and its application across diverse contexts may support stakeholders in understanding the broader impact and transformative potential of these emerging technologies.