Research data scientist and computational social scientist
Qing Chang
Hub for AI and Data Science Leadership (HAIL), University of Pittsburgh
Affiliate: Responsible Data Science, Complexity, and AI for Peace Lab · Resilient Economy Lab
I use AI, data, and rigorous social science methods to generate credible evidence on industrial policy, technological innovation, talent flows, and conflict.
Research areas
Research Toolkit
From Complex Data to Credible Evidence
Research orientation
I work across political science, economics, and data science to study how technological change, especially in green energy and AI, reshapes labor markets and workforce development. My research also asks how evidence-based policy can ease technological disruption while supporting long-run economic adaptation. This work connects substantive social science questions with data-intensive empirical analysis.
Data infrastructure
I regularly build research datasets from administrative records, workforce intelligence data, patents, firm and investment data, and geospatial sources. A large part of my work involves turning complex and messy data into usable research inputs through cleaning, harmonization, linkage, validation, and documentation. These workflows are central to making large-scale data reliable for collaborative research.
Policy inference
My work uses rigorous empirical strategies, including both traditional causal inference tools and newer causal machine learning methods, to study how policy and technological change affect workers, regions, and firms. I place strong emphasis on careful research design so that empirical results remain credible and policy relevant. I have also contributed to methodological tools that make flexible causal analysis more usable for applied social science research.
Research interfaces
I develop interactive tools that help research teams explore data, review results, and use empirical outputs more effectively. This includes building applications that connect backend data workflows with frontend interfaces for research use. I also use AI-assisted development tools to create features that improve research workflows for collaborators and research assistants.
Spatial and relational data
My research often relies on spatial and relational data to study how workers, firms, places, and events are connected. In my work on green investment, I use geocoded investments, local labor market data, and job-to-job flows to study geographic mismatch between workers and emerging opportunities. In my conflict forecasting work, I use spatial and relational structure to represent how actors, places, and events interact over time.
AI research infrastructure
I use knowledge graphs and AI tools to organize complex social and spatial information in ways that support analysis, forecasting, and research applications. In my conflict work, this means structuring messy information so that it can be queried, updated, and used in downstream models and interactive tools. More broadly, this reflects my approach to AI as a way to build stronger research infrastructure rather than simply generate outputs.