Quantitative Methods for Exploring Population Change: perspectives from Epidemiology and Evolutionary Biology applied to Widowhood and Infant Mortality.
Abstract: Changes in the characteristics of individuals and populations are fundamental issues for many disciplines. This talk tries to expand our demographic perspective by illustrating how epidemiologists and evolutionary biologists address similar problems, with a particular focus on selection bias. All disciplines with demography at their core have to recognise that changes for individuals may not be the same as changes for populations. Recuitment and losses to populations, and missing data, are almost always selective on one or more characteristics.
To avoid the complications of model-fitting or hypothesis testing, the two examples come from Exploratory Data Analysis. They require no specialist software and are exact calculations, rather than statistical estimates. The first example illustrates the epidemiological method for attributing a risk to a particular characteristic, in this case widowhood, and assessing its effect on overall mortality. The example of a method from evolutionary biology shows how individual change, recruitment, and losses determine the dynamics of the mean and variance of infant mortality measured at the population level.