Parametric Bayes Analyses to Study the Age-Specific Fertility Patterns

Abstract

Synoptic Abstract: The article considers various parametric models proposed for studying age-specific fertility patterns and develops a Bayesian alternative for their analyses. The analysis is done using weak proper priors for the parameters. Since the posterior analysis involves analytically intractable integrals, the article uses Metropolis algorithm to draw the corresponding posterior samples and hence obtains the sample based inferences. The maximum likelihood estimates of parameters are also obtained by non-linear optimization technique. Such estimates are used as an aid to our Bayesian computation. The considered models are finally compared using different Bayesian model selection tools to recommend an appropriate model for the situation under consideration. Finally, numerical illustration is provided for two different datasets in order to check the feasibility of the models with changing fertility patterns over time. © 2018, © 2018 Taylor & Francis Group, LLC.