Joint Analysis of Repeatedly Observed Mixed Discrete and Continuous Outcomes via Copula Models

• Activity Number: 679
• Type: Topic Contributed
• Date/Time: Thursday, August 8, 2013 : 10:30 AM to 12:20 PM
• Sponsor: Health Policy Statistics Section
• Author(s): Beilei Wu*+ and Alex de Leon
• Companies: PAREXEL International and Univ of Calgary
• Keywords: copula models ; correlated mixed outcomes ; polychoric and polyserial correlations ; non-Gaussian marginal distributions ; pairwise likelihood estimation ; random effects model

Abstract

This presentation is concerned with the analysis of multiple correlated discrete and continuous outcomes that are observed on the same subject over time. Joint analysis of such disparate responses (i.e., mixed discrete and continuous outcomes) is problematic in practice due mainly to the difficulty of defining a joint model. Our proposed approach is based on a random effects model that accounts for associations between the outcomes (of the same or of different types) for the same subject at the same time point, and/or at different time points. A latent-variable approach is adopted to sidestep complications of direct application of copula models to discrete data. The approach yields regression parameters that are marginally meaningful, and permits the adoption of flexible non-Gaussian marginal distributions for the mixed outcomes as well as for the random effects. Full and pairwise likelihood estimation methods are implemented for the model using PROC NLMIXED in SAS. The proposed methodology is illustrated using individual panel data on the wages, work hours, and union memberships of 545 men during the period 1980-1987 (Vella and Verbeek, 1998).