Date of Submission



The second wave feminist movement during the mid to late 20th century saw rapid advancements in contraceptive, legal, and economic rights of women. However, despite recent advances in women’s liberation, gender differences in educational outcomes (e.g., self-beliefs, attitudes, aspirations and educational attainment in science, technology, engineering and mathematics [STEM]) remain heavily entrenched. This thesis explores the notion that the gender gap in STEM has its origins in self-beliefs and task values of young people. However, less is known about how much of the STEM gender gap can be explained by these Expectancy Value Theory constructs. Moreover, there is a lack of research that utilises an intersectional lens to explore how social and cultural context moderates the size of gender gaps in self-beliefs and attitudes towards math and science. This thesis addresses these research gaps utilising meta-analytic, longitudinal, and interview data. Meta-analytic findings from 176 studies in Study 1 show that gender differences in expectancy value constructs are domain specific, and that there are significant moderation effects across social class, gender equality, and gender segregation in university enrollments. Study 2 explores the replicability of meta-analysis results from Study 1, and extends upon these results through an analysis of a large nationally representative database (n = 10,370) that includes ethnicity, geography, and educational attainment. Using the same database, results from Study 3 show that while EVT can account for some of the gender disparity in STEM enrollment, there is still a very large amount of difference that remains unexplained by current theory. Furthermore, results indicate that even when comparing male and female students of equal ability and attitudes, young women still are significantly disadvantaged in terms of STEM university enrollment. A content analysis in Study 4 explored whether open-ended interview data from young Australians who enrolled in a university STEM course (n = 447) versus those who chose to discontinue their STEM education after senior high school (n = 949) can add to current theory. Results point the role of dimensional comparison as critical to educational choices, but again, there were no major themes that arose that significantly deviate from current theory. Results are discussed in light of future directions for research, and implications for policymakers and educators.


Institute for Positive Psychology and Education

Document Type


Access Rights

Open Access


311 pages

Degree Name

Doctor of Philosophy (PhD)


Faculty of Health Sciences