Precollege science achievement growth: Racial-ethnic and gender differences in cognitive and psychosocial constructs

The purpose of this study was to gain a more complete understanding of the differences in science, mathematics and engineering education among racial-ethnic and gender subgroups by exploring factors related to precollege science achievement growth rates. Using Hierarchical Linear Modeling (HLM) and multi-wave, longitudinal data from the first three waves of the National Education Longitudinal Study of 1988--1994 (NELS:88/94), this study examined precollege science achievement growth rates during the 8th to 10th grade period and the 10th to 12th grade period for African American males, African American females, Latino males, Latina females, Asian American males, Asian American females, White males and White females.;For the 8th--10th grade period, previous grades were significantly and positively related to science achievement growth for all subgroups; and socio-economic status and high school program were significantly and positively related to science achievement growth for all subgroups except one (Latino males, and Asian American males respectively). For the 10th--12th grade period, the quantity of science courses completed (science units) was the only variable that was statistically significant for more than one racial-ethnic by gender subgroup. Science units taken were significantly and positively related to 10 th--12th grade growth rates for all racial-ethnic by gender subgroups except Latino males.;Locus-of-control was the only cognitive or psychosocial factor included from Eccles, Adler, Futterman, Goff, Kaczala, Meece and Midgley's (1983) theoretical framework for achievement behaviors that appeared to exhibit any pattern across race-ethnicities. Locus-of-control was positively related to 8th--10 th grade science achievement growth for females across all racial-ethnic subgroups, as well as for African American males. However, for both the 8 th--10th grade and 10th--12 th grade periods, there was no consistency across racial-ethnic or gender subgroups in effects of the remaining cognitive and psychosocial factors on science achievement growth rates. Cognitive and psychosocial variables were statistically significant only for specific subgroups, and did not generally exhibit any commonalities across gender, or by race. The findings accentuated the importance of disaggregating data and analyses by both race-ethnicity and gender.