Bayesian latent-class mixed-effect hybrid models for dyadic longitudinal data with non-ignorable dropouts.

ABSTRACT: The analysis of longitudinal dyadic data is challenging due to the complicated correlations within and between dyads, as well as possibly non-ignorable dropouts. Based on a mixed-effects hybrid model, we propose an approach to analyze longitudinal dyadic data with non-ignorable dropouts. We factorize the joint distribution of the measurement and dropout processes into three components: the marginal distribution of random effects, the conditional distribution of the dropout process given the random effects, and the conditional distribution of the measurement process given the random effects and missing data patterns. We model the conditional dropout process using a discrete survival model, and the conditional measurement process using a latent-class pattern-mixture model. These models accou

SUBMITTER: Ahn J 

PROVIDER: S-EPMC3970927 | biostudies-literature | 2013 Dec

REPOSITORIES: biostudies-literature