– The mean SES of the school had a significant effect on students’ mathematics course taking for African American, Hispanic, and White students, when controlling for family and academic backgrounds of the students attending the school.
– A high schools’ racial composition was not a significant factor except for Asian students on mathematic course-taking.
– The negative effect of percentage high minority schools can be explained by school mean SES.
– Gender and racial/ethnic differences are large in the mathematics course-taking with females and ethnic minorities, with the exception of Asian Americans, taking fewer advanced mathematics courses.
– After controlling for students’ SES and previous achievement- students’ mathematics course-taking and mathematics attitudes and behaviors were significant predictors of their college major choice.
– African American and Hispanic students were more likely to take advanced mathematics courses if they attended schools where students perceived a higher academic climate, teaching quality, and parental participation.
– Mathematics competency was a significant predictor for White female students’ STEM major choice.
– Mathematics affection had a positive effect on Hispanic and White students, especially for female students on their advanced mathematics course-taking.
– The results of this study suggest that gender disparities in STEM choice occur because girls are less likely to pursue STEM majors, but racial/ethnic disparities occur due to the underachievement of African American and Hispanic students in their advanced mathematics participation in high school.
2013 - Gender and ethnic differences in precollege mathematics course work related to science, technology, engineering, and mathematics (STEM) pathways
The conceptual framework used theorizes that student outcome (i.e., mathematics courses-taking) is the result of student-level factors as well as school-level factors. The framework also theorizes reciprocal relationships among these factors in the two levels. More specifically, at the student level, academic course-taking is a function of the effects of students’ background characteristics influencing student behavior and attitudes, which then is related to advanced math course-taking. For school-level influences, school context impacts school process, which ultimately factors into advanced math course-taking. Because students-level factors are nested within school-level factors, there is also reciprocity of effects in which student characteristics can influence school characteristics.
Education Longitudinal Study of 2002. Based on the data from approximately 12,160 students from 752 public schools. The second analysis that examines the predictions of STEM majors used a subset of 10,599 students who completed surveys in 2006.
The conceptual framework theorizes that student outcome (i.e., mathematics courses-taking) is the result of student-level factors as well as school-level factors. However, there is a possibility that students who attend the same school do not take the same courses or display the same course-taking patterns. Thus, in addition to examining school factors, this study also explored the student factors that may influence students’ course-taking in high school and the choice of STEM major in postsecondary institutions.
The independent variables were background variables (SES variable was measured by father’s education level, mother’s education level, father’s occupation, mother’s occupation, and family income), early academic performance, mathematics attitudes and behaviors, college preparatory program (used as a control variable), parents’ educational expectations (used as a control variable), mean school SES, percent minority, student-teacher ratio, and school process (includes the percent of students in college preparatory program, students’ mean perception of academic press, and mean parent participation).
The dependent variable was high school course-taking. The analysis utilized in this study used a condensed version of Burkam and Lee’s (2003) eight-level measure where the lowest categories (no, non-academic, and low academic) were combined to yield sufficient cell sizes for statistical estimation.