GenAI Meets Psychometrics: Development and Validation of the Human-AI Collaboration Dynamics Scale and the Generative AI-Research Augmentation Scale

This study develops and validates two psychometric scales, namely Human-AI Collaboration Dynamics (HACD) and Generative AI-Research Augmentation (GAI-RA), and examines the role of human-AI collaboration in scale development itself. Data were collected from university staff, doctoral students, master’s students, and research-active undergraduates from 16 English Medium Instruction (EMI) universities in Mainland China, Hong Kong, and Macau. Exploratory factor analysis on one subsample (N_A1 = 344) identified underlying dimensions, followed by confirmatory factor analysis on two independent subsamples (N_A2 = 344; N_B = 447). The HACD scales produced four versions, from a core two-factor to an expanded five-factor structure, providing a flexible measurement toolkit. The GAI-RA scale emerged as a concise, single-factor, five-item measure of GenAI research augmentation. These validated instruments, with excellent psychometric properties, advance the study of human-AI collaboration in academic research and demonstrate that GenAI can be integrated into a psychometrically rigorous scale development and validation process.