About Me
I am a Ph.D. candidate in Sociology and Demography at Pennsylvania State University. Before entering Penn State, I received my master’s degree in Economics from University of California, Los Angeles and Bachelor’s degree in Economics from Peking University.
My Chinese name is 徐家汇, exactly the district name in Shanghai, China. Yet, I am not from Shanghai and my hometown is Henan province.
I am on the 2024-2025 job market.
My email address is jiahuixu [at] psu [dot] com.
Research
I am a quantitative and computational sociologist with an empirical emphasis on social stratification and inequality research.
My research is organized around three aspects of social stratification and inequalities: (1) racial and ethnic health disparities over the life course; (2) economic disparities across social groups; and (3) innovative quantitative and computational methods and open-source statistical packages. I develop and apply quantitative methods to examine the prevalence and dynamic nature of inequalities in health, education, poverty, occupation, and earnings, utilizing large-scale surveys and vast administrative data containing over two billion person-year observations. For example, I examine how social institutions stratify the health outcomes in U.S. and how exposure to the social safety net, as well as education, mitigate the inequalities. I also study how access to and completion of higher education both equalize and stratify life chances over the life course in different cohorts. Moreover, I assess how occupations explain the rise of earning inequalities in the United States.
I am interested in both demographic methods and machine learning methods. I develop new methods at the intersections of machine learning, causal inference, and moderated mediation. I also develop open-source statistical packages to implement these methods in estimating causal effect heterogeneity with machine learning techniques, analyzing age-period-cohort effects for nonlinear decompositions, and assessing occupational write-in variations using natural language models. In my latest work, I develop machine learning algorithms to faciliate multiply robust causal inference (see my lead-author working paper “Flexibly Detecting Effect Heterogeneity with an Application to the Effects of College in Reducing Poverty” and my dissertation chapter 1). I also use text mining techniques to collect data from websites to study the unintended effects of social institutional changes. Additionally, I fine-tuned and deployed large language models to automatically code occupational write-in data (try the autocoder).
My previous work have been published on Sociological Methodology, SSM-Population Health, BMJ Open, R Journal, Scottish Journal of Political Economy.
I was elected to a two-year term as a Student Representative for the ASA Methodology Section in 2024.
Check out Research for details.
Software
I like programming and I wrote several R packages on CRAN: