Metrics and Models is a Health and Social Data Science Lab which runs an open seminar series: all are welcome to attend.
Thursday, 28 May 2026, 2pm to 3pm
Talk Title: Using AI Predictions to Augment Rather Than Replacing Surveys
Abstract: Precision in household surveys is increasingly costly to buy with additional interviews, yet many statistics that inform health, labor, and social policy still require narrow confidence intervals to be useful. Our question is whether we can obtain some of that precision “for free” by exploiting information already available at population scale. Prediction‑Powered Inference (PPI) provides principled framework to achieve this: obtain a predictor using either existing data or external information (LLM), predict it on a very large auxiliary file, and then calibrate the prediction‑based estimate with a gold‑standard correction computed on the survey labels. As a demonstration, we apply PPI to National Health Interview Survey (NHIS) and General Social Survey (GSS), harmonized with the American Community Survey (ACS), covering a broad range of outcomes that include political opinion, socio-economic conditions, and health conditions and behaviours. We show that the combination of auxiliary population-level information and external prediction algorithms can increase the effective sample size of survey data.
Bio: Lai Wei is an Assistant Professor of Sociology and HKU-100 Scholar at the University of Hong Kong. Prior to his post he obtained his PhD in sociology from Princeton University. He studies social stratification and quantitative methodology. His past works have been published in Sociological Methods & Research, Journal of Royal Statistical Society, Journal of Health and Social Behaviour, among other outlets.
Series: Metrics and Models
Department: Nuffield Department of Population Health (Department)
Organiser: Metrics and Models
Host: Metrics and Models