The Mapping Job Structure (MapJob) Project develops a relational framework to study the structure and evoluation of jobs. This framework highlights the inter-connectedness between jobs, workers and their associated attributes (e.g. skills, location). This project combines various sources of data from the Census Bureau, Bureau of Labor Statistics, ONET, as well as data gathered from online job postings and resumes. This project received support from the Russell Sage Foundation and the Washington Center for Equitable Growth.
The Communities and Connections (C&C) Project combines large-scale datasets on human mobility with Census and administrative data to examine inequality in individuals’ and communities’ connectedness to resources and opportunities. It also explores how community-level interventions may help reduce this inequality. Connectedness to a network of places, facilities, social circles, local institutions, and job opportunities (or lack thereof) shape many aspects of youth’s life chances including educational attainment, upward social mobility, economic opportunities, physical and mental health, and political participation. While my prior research has mostly considered individuals and communities as independent units of analysis, this project will look into the complex network of connections between them. This project is supported by the William T. Grant Scholars Program.
The Social Dynamics through Images Project aims to build a large-scale, de-identified data infrastructure comprising time-stamped, geo-tagged photographs of public spaces. We then train machine learning and AI-based vision models to detect "social encounter groups" in these images—ranging from people dining together in a restaurant to people sharing a library table. We combine automated coding with human coding to characterize individuals in these social encounters and study their social dynamics. This dataset will provide the first opportunity to understand the magnitude, sources, and outcomes of micro intergroup exposure.
The Inter- and intra-generational (Inter-intra) Mobility Project builds on recent developments in machine learning and causal inference to understand how inequality unfolds over the life course and across generations.
The Pathways to Publication (Path2Pub) Project digitizes two decades of social science conference programs and uses machine learning algorithms to match them to a comprehensive bibliography of academic publications, aiming to understand how social inequality translates into publication bias concerning certain groups, research topics, and author charateristics.