Benefits of Integrated Population Modeling

Abstract

The joint analysis of data sets from population and individual levels through an integrated population model provides a number of benefits highlighted here. Key benefits of integrated population models (IPMs) are the increased precision of parameter estimates and the ability to estimate parameters without explicit data—the so-called hidden parameters. Both features are the result of efficient exploitation of information from population-level data. Information about population size and demographic rates originating from each component data set in the IPM is averaged, and by cloning component data sets, we can “measure” the flow of information among data sets. Further benefits of IPMs are the ability to estimate correlation among demographic rates (cross-correlations) and population structure. IPMs offer much flexibility, allowing for creative combining of data sets that have been collected using disparate sampling protocols. There is also freedom in choosing statistical distributions for the observation model of the population count data, often with only minimal effects on results. Despite the multiple benefits that IPMs offer and the high flexibility they provide in terms of data types that can be included, we advocate care—estimates must be critically checked (in terms of parameter identifiability and sensitivity to priors) before they can be used for inference.