Pieces of the immigrant paradox puzzle: Measurement, level, and predictive differences in precursors to academic achievement

The "immigrant paradox" indicates that the academic attitudes and outcomes of 1st-generation youth exceed those of the 2nd- and 3rd-generation. This paper examines a) whether unobserved measurement bias contributes to these generational differences, b) generational differences in levels of behavioral school engagement (BSE) and perceived supportive school relations (SSR), and c) to what extent BSE mediates the relations between SSR and academic achievement and whether these relations differ across generations. New York City Social and Academic Engagement Study (NYCASES) data were analyzed. Strong measurement invariance for BSE and SSR suggests that unobserved measurement bias does not contribute to the immigrant paradox. 1st generation youth evinced higher latent means for BSE and SSR than 2nd or 3rd-generation youth. 1st generation youth responded to SSR by exerting effort while 2nd and 3rd generation youth responded to SSR by complying with rules. Because effort engendered achievement more than compliance, this study identifies a mediating mechanism that contributes to the immigrant paradox.