Age at First Marriage in Malawi: a Bayesian multilevel analysis using a discrete-time model

This paper presents a hierarchical discrete-time survival model for the analysis of data on timing of marriage. The data were collected as part of the Knowledge, Attitudes, and Practices in Health (MKAPH) survey for Malawi held in 1996. The model assesses the determinants of transition to marriage and accounts for family and community random effects with cross-level correlation structure. A nonparametric technique is used to dynamically model the baseline discrete hazard. The model produces improved estimates of fixed effects and the associated standard errors, in addition to providing intracluster and intercluster correlations at the levels of the hierarchy. A full Bayesian approach is employed using Markov Chain Monte Carlo for posterior computation.