Development and validation of the AI cognitive offloading scale for Chinese college students

Introduction Against the backdrop of rapid advances in generative artificial intelligence (AI), college students increasingly delegate cognitive tasks to AI systems. This study developed and validated the AI Cognitive Offloading Scale (AI-COS) as an alternative, context-specific operationalization for Chinese college students, focusing on the cognitive functions delegated to AI during learning. Methods An initial pool of 26 items was generated through literature review, theoretical analysis, and expert evaluation. Study 1 used data from 665 college students for item analysis and exploratory factor analysis. Study 2 used an independent sample of 1,123 college students for confirmatory factor analysis, reliability testing, criterion-related validity assessment, and measurement invariance testing across gender. Results Exploratory factor analysis yielded a 20-item scale with three dimensions: memory offloading, assistive offloading, and thinking offloading. The second-order three-factor model showed satisfactory fit (CFI = 0.951, TLI = 0.944, RMSEA = 0.047, SRMR = 0.036). The scale demonstrated strong internal consistency (Cronbach's α = 0.944), split-half reliability (0.932), and test-retest reliability (0.937). Convergent validity was satisfactory, and full scalar invariance across gender was reported. The total score and all three dimensions were positively associated with AI dependence and negatively associated with metacognitive monitoring. The total score and assistive offloading were negatively associated with need for cognition, whereas memory and thinking offloading were not significantly associated with need for cognition. Discussion The findings provide initial evidence for the reliability and validity of the AI-COS. The three dimensions represent distinguishable functions of cognitive delegation rather than a developmental sequence. The scale offers a context-specific instrument for research on AI-mediated cognition, learning regulation, and student development.