Psychometric evaluation of the academic involution scale for college students in China: An application of Rasch analysis

Recent academic attention on educational involution in China underpins the need for a valid and reliable instrument to precisely measure college students' academic involution behaviors. Seeing the scarcity of a proper instrument, the current study attempted to analyze the item-level psychometric properties of the newly developed Academic Involution Scale for College Students (AISCS) in China by using a Rasch model. A total of 637 college students in a public university in northern China participated in the study. Data were examined with respect to unidimensionality, rating scale functioning, item fit statistics, item polarity, item- and person-level reliability and separation, item hierarchy and invariance across educational background with Winsteps. The results show that AISCS was a single unidimensional construct with good psychometric properties. Although two items demonstrated differential item functioning, it is plausible given the differences between assessment methods for undergraduates and postgraduates. Limitations and future research directions with regard to sample selection, inclusion of more validity evidence and adding prospective additional academic involution were discussed.