January 13, 2016 – January 20, 2016
Max Planck Institute for Demographic Research, Rostock, Germany
Population projection matrices describe the dynamics of individuals, cohorts, and populations, in a way that generalizes traditional life table methods. This course will introduce the use of matrix methods to analyze some of the basic problems of health demography: measures of longevity and healthy longevity, the dynamics of individual transitions among health and disease states, and the projection of the health composition of populations. The matrix approach will be used to extend the typical analyses in two directions: (1) the calculation of average quantities will be accompanied by analyses of variances due to individual stochasticity, and (2) sensitivity analyses will be introduced to evaluate the effects of parameters on the demographic results. Methods to be introduced include Markov chains, Markov chains with rewards, multistate matrix models, and matrix calculus methods for sensitivity analysis. These models will be compared to some of the traditional approaches to health demography.
The faculty for this course: Hal Caswell and Neil Mehta.