PREDICTIVE VALIDITY OF STUDENT AND SCHOOL VARIABLES: A MULTIPLE REGRESSION ANALYSIS ON DIVERSE STUDENTS’ MATHEMATICS ACHIEVEMENT

Abstract

This study explored the relationship between race and math achievement, considering gender and the moderating effect of socioeconomic status (SES) in a diverse high school student sample (n=200) from the National Education Longitudinal Studies (NELS) program. Surprisingly, after controlling for gender, Hispanic students outperformed the reference group (White) in math scores, challenging previous assumptions. However, no significant differences were found for Asian and African American students. Contrary to existing literature, SES did not moderate the race-math achievement link. Similarly, when controlling gender, race, and SES, school type did not significantly affect math achievement. Among predictor variables, program type emerged as the most influential, highlighting the importance of educational programs in boosting math skills. Gender and school type had limited predictive power in determining math performance. In summary, this study exposes race-related math achievement disparities in diverse student populations and highlights the limited role of SES and school type in moderating these gaps. These findings provide valuable insights for educators and policymakers striving to address math achievement disparities and emphasize the necessity for further research into this intricate interplay of factors.