I serve as the technical and methodological lead for Data Labs at Pew Research Center, where we study how technology shapes public opinion and behavior. My work sits at the intersection of computational methods and social science, combining digital trace data (web browsing, platform APIs, social media) with representative survey research to answer questions neither approach can address alone.

Recent work includes measuring how AI-generated search summaries change real-world click behavior, analyzing what Americans’ web browsing data reveals about AI exposure, and studying following and engagement patterns on TikTok and Reddit. I’m currently focused on how LLMs can extend what’s possible in public opinion research, including active work evaluating LLMs as synthetic survey respondents.

I hold a Ph.D. in Political Science from the University of California, Riverside. Before joining Data Labs, I was a researcher at AAPI Data, where I worked on public opinion research on Asian American and Latino communities for nearly a decade.

My email address is sonoshah [at] gmail.com