Estimating vital rates in the developing world: A Bayesian process modeling approach

Though quality data on basic population indicators such as births and deaths are vital for forming and evaluating policy and public health, only a small subset of the world's countries maintain ongoing, full coverage civil registration systems. This proposal contributes statistical methods to combine demographic data which arise from multiple sources, using differing sampling frames, and are of highly variable quality. The proposed methods leverage lessons learned from surveys and demographic surveillance systems to develop new strategies for estimating temporal trends in fertility and quantify non-sampling errors that present major obstacles to obtaining reliable fertility rates from surveys.