Date of Award

5-1-2022

Language

English

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

College/School/Department

Department of Educational and Counseling Psychology

Program

Educational Psychology and Methodology

Content Description

1 online resource (vi, 179 pages) : illustrations (some color)

Dissertation/Thesis Chair

Mariola Moeyaert

Dissertation/Thesis Co-Chair

Hayword Horton

Committee Members

Aaron Benavot

Keywords

Gender Equity, Girls Education, Hierarchical Linear Modeling, International Education, Large-scale Data, TIMSS, Mathematics, Academic achievement, Discrimination in education, Educational equalization, Girls

Subject Categories

Educational Psychology | Education Policy | Science and Mathematics Education

Abstract

Using the Trends in International Mathematics and Science Study Data (TIMSS) 2015 dataset, this study examines 30 different contextual indicators to determine significant predictors of girls’ mathematics achievement globally. The study design employs three nested levels in the hierarchical linear model (individual, classroom, and nation) to analyze cross-national scores and responses to the contextual questionnaires. Additionally, the focus is on girls as a standalone, independent population, not in comparison to boys. This research seeks to understand at which level of society the most variability is found, as well as analyze the comparative effect sizes of various explanatory contextual predictors within the model. By assessing specific aspects of girls’ education at the individual (household), classroom and national level, and determining at which level the most variability occurs, the model clarifies the efficacy of different policy approaches. The study found the third (national) level explains an unexpectedly high amount of the variability in girl’s mathematics achievement. Additionally, the patterns found at all three levels in this model more closely adhered to smaller, single cohort research focusing on gender differences, than to previous research using large-scale mixed gender datasets.

Share

COinS