Construct Validity, Longitudinal Measurement Invariance, Incremental Validity, and Predictive Validity of the Original Grit Scale in Chinese Young Adults

Although many studies have attempted to validate grit scales because of the construct's popularity, most have considered the shorter rather than the longer Original Grit Scale (Grit-O). We examined the Grit-O's construct validity, longitudinal measurement invariance, incremental validity for academic performance, and longitudinal predictive validity for subjective well-being among young Chinese. We used a cross-sectional sample of 3,322 college students and a longitudinal sample of 1,884 college students, tested twice over 10 months. The first-order factor model fit the data better than other models and showed partial configural and metric measurement invariance over time. Grit and its two facets longitudinally predicted subjective well-being (i.e., life satisfaction, happiness, positive affect, negative affect, and depression) but had negligible incremental validity for two semesters' grades after controlling for conscientiousness. So, while the Grit-O could be a useful construct for young adults, its predictive value overlaps with a better-established construct, conscientiousness.