Gender and Racial Inequalities in School-to-Work Linkage: The Interplay Between Horizontal Mismatch and Educational Field Specificity
Institute for Diversity Science Seed Grant, 2024-25
Gender and racial wage inequalities persist and have even expanded among highly educated workers, despite remarkable educational gains of women and minorities. Extant research addresses both educational and occupational factors in explaining these disparities. However, an overlooked potential source of inequality lies in the school-to-work linkage: the horizontal mismatch between educational and occupational fields. Women and racial/ethnic minorities may be disproportionately relegated to mismatched jobs and face greater penalties when mismatches occur; these disparities can also depend on educational field specificity, defined as the extent to which an educational field is linked to a few occupations. Utilizing nationally representative data from the American Community Survey on over 2.5 million college graduates and the National Survey of College Graduates on over 124,000 individuals, I investigate: 1) Are there gender and racial disparities in the prevalence and consequences of horizontal mismatch? 2) Do these disparities vary according to levels of educational field specificity? I replace previously-used, crude dichotomous definitions of horizontal mismatch and educational field specificity with objective data-driven measures to capture these complex concepts on continuous spectrums. Findings will help guide college major choice, career services, and policies and initiatives to address inequality in school-to-work linkages.
The Transformative Potential of Generative Artificial Intelligence (GAI) in STEM Learning, Equality, and Inclusion: A Student-Centered Mixed Methods Study
Institute for Diversity Science Seed Grant, 2023-24
The rapid development and widespread adoption of Generative Artificial Intelligence (GAI) tools, such as ChatGPT, necessitate an urgent examination of their implications for education and inequality. GAI may provide personalized, low-expense, accessible, and culturally diverse academic support to historically marginalized and underrepresented students, potentially improving diversity and inclusion in higher education. However, concerns have been raised regarding access disparities, academic integrity, and potential bias in GAI-generated content that may perpetuate inequality. Adopting a constructively critical perspective, this study examines the role of GAI in shaping college students’ learning experiences in STEM fields and its potential impact on gender, racial/ethnic, and socioeconomic inequalities. It aims to answer: (1) How do college students utilize GAI to facilitate STEM learning? What are the barriers, facilitators, and other influencing factors that shape their experiences with GAI across socio-demographic groups? (2) How does GAI perpetuate or alleviate existing inequalities? Are there variations in GAI acceptance, attitudes, usage patterns, and concerns among students from different socio-demographic backgrounds? How may historically underrepresented students leverage GAI to overcome challenges in STEM learning? (3) What institutional support and policies are necessary to guarantee an inclusive learning environment with GAI? Employing a student-centered, transformative sequential mixed methods approach, this study leverages the strengths of large-scale quantitative data from a campus-wide survey, fine-grained qualitative insights from targeted in-depth interviews, and innovative textual data from real student-AI conversations. Findings will inform the development of inclusive AI-enhanced learning strategies and policies to support diverse student populations.
Social Connectedness, Network Embedded Inequality, and Disparities in Education Outcomes
AERA-NSF Research Grant, 2022-23
This project is funded by a grant from the American Educational Research Association (AERA) which receives funds from the National Science Foundation. Using big social networking data, it aims to develop three sets of measures of network characteristics at the county level: 1) network heterogeneity; 2) network-embedded socioeconomic resources; and 3) network-embedded inequality structures. Linking these measures to county-level data on student academic performance and disparities, this project will examine four research questions: 1) Are counties with similar levels of socioeconomic and education resources more connected, thus reinforcing inequality of resource distribution? 2) How is county network heterogeneity associated with county average student academic performance and performance disparities across class, gender, and racial groups? 3) How are county embedded social network resources associated with county average student academic performance and disparities? 4) How are county embedded network inequality structures associated with county average student performance disparities?