New Scale Measures Student Generative AI Reliance in Writing.

As generative AI (GenAI) becomes increasingly integrated into academic writing, understanding how students use these tools, rather than just if they use them, is crucial for learning, academic integrity, and equity. Existing measures of AI reliance have limitations, often developed inductively and validated with narrow samples. This research addresses that gap by developing and validating the Generative AI Reliance Types Scale (GenAI-RTS). The GenAI-RTS is a 20-item instrument designed to measure four theoretically derived types of GenAI reliance: Strategic, Instrumental, Dependent, and Dialogic. Through a multi-source validation framework, including a survey of 382 undergraduates and interviews with 14 students, the study confirmed a five-factor structure where Strategic Reliance splits into "Deliberate Use" and "Critical Evaluation." The scale demonstrated acceptable to good reliability and scalar measurement invariance across various student demographics. The findings indicate that strategic reliance correlates positively with AI literacy, and the different reliance types distinguish students across various writing process and outcome variables. This new scale offers a robust, theoretically grounded tool for educators and researchers to identify student reliance profiles and inform AI literacy interventions.