Nonparametric Bayes Models for Mixed-Scale Longitudinal Surveys

• Activity Number: 619
• Type: Invited
• Date/Time: Thursday, August 13, 2015 : 8:30 AM to 10:20 AM
• Sponsor: IMS
• Author(s): Tsuyoshi Kunihama* and Amy Herring and Carolyn Halpern and David Dunson
• Companies: Duke University and The University of North Carolina at Chapel Hill and The University of North Carolina and Duke University
• Keywords: Mixed-scale data ; Mixture model ; Multivariate longitudinal data ; Nonparametric Bayes ; Sample survey ; Stratified sampling

Abstract

In many social science surveys, participants repeatedly respond to questions over time. Modeling and computation for multivariate longitudinal data has proven challenging, particular when data are not all continuous and Gaussian but contain discrete measurements. Also, study participants are often selected via stratified random sampling, leading to discrepancies between sample and population. To adjust for this gap, survey weights are constructed but it is not clear how to include them in hierarchical models. Motivated by an application to sexual preference data, we propose a novel nonparametric approach for mixed-scale longitudinal data in surveys. The proposed approach relies on an underlying variable mixture model, with time-varying latent factors. Bias from survey design is adjusted for in posterior computation relying on a Markov chain Monte Carlo algorithm. The approach is assessed in simulation studies, and applied to the National Longitudinal Study of Youth.