School Variability and Associated Factors in Within-school Gender Differences in Mathematics Performance
- Author(s): Yun, Jung-eun Ellie
- Advisor(s): Webb, Noreen M
- Seltzer, Michael
- et al.
Over the past few decades, education researchers have repeatedly found that male students outperform their female peers on standardized mathematics tests. The attention has turned to factors associated with this phenomenon, with student characteristics and social contexts being a prime focus. Less studied, however, is variability in the magnitude of gender differences in mathematics performance across schools and the impact that school environments might play in this gender difference. The current study seeks to: a) investigate whether differential gender performance exists in a large-scale mathematics assessment and to what extent gender differences vary across schools and b) identify school factors associated with school variability in the size of gender differences in mathematics performance.
The current study uses data from the Trends in International Mathematics and Science Study (TIMSS) collected in 2015 from 8th graders and high performing 12th graders in the US. Hierarchical Linear Modeling (HLM) is the main statistical analysis framework for the study. In particular, this study uses a two-level HLM models with students as the level I model unit and schools or teachers as the level II model unit.
The results indicate that for students in both 8th and 12th grades, males perform better than females. There was also substantial variability between schools in the size of the gender difference in mathematics performance. School-level HLM analyses indicate several predictors are associated with the within-school gender gap, such as the presence of a mathematics resource shortage, communication among teachers, teacher confidence, and teacher support of student participation. For high performing12th graders, whether teachers felt their school was safe and orderly was associated with the within-school gender gap in mathematics performance.
Further analyses using a teacher-level HLM demonstrate that female teachers and teachers under age 30 are more likely to have classes with a wider gender gap in mathematics performance. In addition, and consistent with results from the school-level HLM, teachers who support student participation are more likely to have classes with a smaller gender gap in mathematics performance. Altogether, this study shows the complexity of the within-school gender gap in mathematics performance and suggests the need for future studies.