Bayesian demography made (almost) simple Seminar

Abstract: Demographers are beginning to embrace Bayesian statistical methods. However, the barriers to entry are high. To tackle realistic problems, in research or in practical applications, demographers need to learn significant amounts of mathematical statistics and statistical computing. The talk will look at an attempt to lower the barriers. We are trying to develop a small suite of models that can be applied to a wide variety of tasks in demography, and to implement them in user-friendly R packages. One of the key challenges is developing sensible defaults, which requires us to identify good demographic or statistical practice, and turn it into computer code. Another challenge is to develop models that are easy to understand, but that perform adequately on a wide range of problems.