AI-Augmented Survey Research: Promises, Pitfalls & Eme(R)ging Tools- Soubhik Barari- nyhackr Oct MU

Visit https://www.nyhackr.org to learn more and follow https://x.com/nyhackr Talk Title- AI-Augmented Survey Research: Promises, Pitfalls, and Eme(R)ging Tools Talk Description- Can large language models (LLMs) improve the way we design, run, and analyze opinion surveys? This talk looks at both the upsides and the risks. I’ll share results from evaluations conducted by research methodologists at NORC at the University of Chicago and the broader research literature, showing how LLMs can act as both “interviewers” and “interpreters” in the opinion survey process. Along the way, we’ll talk about what actually works, what doesn’t, and what that means for the analysis of opinion and behavioral data more broadly. We’ll wrap up with a quick tour of new R tools for integrating LLMs into survey data workflows. Bio- Soubhik Barari is a quantitative social scientist and data scientist. Currently he is a Research Methodologist at NORC at the University of Chicago, working on applied problems at the intersection of survey methodology and data science. He is also an Adjunct Associate Professor of Political Science at Columbia University where he teaches data science, elections, and political communication. Previously, he's worked in research and data science roles at SurveyMonkey and Microsoft Research.