Stereotype threat and college academic performance: A latent variables approach
This paper investigated how multiple indicators of stereotype threat are associated with college academic performance using data from the National Longitudinal Survey of Freshmen (NLSF) (N = 2,565; 918 Blacks, 794 Hispanics, 853 Whites). Academic performance, measured by average GPA in the freshman/spring and sophomore/fall semesters, were related to measures of internalization (believing that valid negative stereotypes are self-descriptive), externalization (believing that invalid negative stereotypes are held by others), academic effort (interest and effort focused on academic performance), and academic performance burden (believing that ones academic performance reflects upon ones group). Findings from structural equation modeling (SEM) showed that externalization (common under stereotype threat) is associated with an increased academic performance burden for Hispanic and Black students. Moreover, this extra psychological burden was associated with poorer academic performance. Internalization was associated with a reduction in academic effort, which also predicted lower academic performance. Controlling for academic effort, however, produced a positive relationship between internalization and academic performance. This suggests that reduced academic effort plays a critical role linking internalization and poor academic performance. Findings also showed that externalization/internalization mechanisms were irrelevant for whites performance. Additionally, the two mechanisms did not appear to be distinct concepts for White students. These data attest to the importance of variables relevant to stereotype threat in real-world academic performance.