<![CDATA[This study aims to adapt and validate the ChatGPT Usage Scale among Indonesian university students. Using a quantitative psychometric approach, data were collected from 542 active university students across Indonesia selected through convenience sampling. The adaptation followed the International Test Commission (2017) guidelines. Content validity using Aiken’s V showed a high value (0.894). The initial Confirmatory Factor Analysis (CFA) indicated poor fit (CFI = 0.741; TLI = 0.698), but after removing eight items with factor loadings below 0.40, the seven-item model achieved good fit (CFI = 0.949; TLI = 0.924; RMSEA = 0.056; SRMR = 0.034). The Cronbach’s Alpha coefficient (0.703) indicated acceptable reliability. Findings revealed that the original three-dimensions structure (Academic Writing Aid, Academic Task Support, and Reliance and Trust) was not fully supported in the Indonesian student context. The Reliance and Trust dimension could not be retained due to insufficient valid indicators. Despite achieving a statistically fit model, the adapted version remains conceptually limited and requires further refinement to comprehensively capture ChatGPT usage behavior among Indonesian students. Indonesian students primarily use ChatGPT instrumentally with limited trust, highlighting the need for higher education institutions to integrate AI literacy, ethical usage guidelines, and reflective learning practices. Furthermore, the adapted scale provides a foundation for future research to better understand AI utilization in academic contexts.]]>
Copyrights © 2025
