Development and validation of a scale of assessing students' integration of generative AI in learning based on the SAMR model

Abstract

This study aimed to develop and validate a scale for measuring students' integration of generative AI, grounded in the SAMR (Substitution, Augmentation, Modification, and Redefinition) model. The study included 1295 participants selected through convenience sampling. Item analysis was first conducted to examine the discrimination and reliability of each item. Exploratory factor analysis results identified a four-factor structure corresponding to the SAMR model's dimensions. Moreover, the Confirmatory Factor Analysis results supported this structure and further revealed a second-order factor structure that included two higher-order dimensions: Enhancement and Transformation, which represent the broader scope of generative AI integration in education. This scale provides a psychometrically sound tool for assessing students' integration of generative AI in learning, offering valuable insights for future research. It demonstrates that the SAMR model is an effective framework for differentiating the various levels at which students utilize generative AI, informing instructional design and strategies for enhancing AI integration in educational contexts.