ABSTRACT
This chapter presents the main approaches to optimization of adaptive survey designs (ASD) and uses the Interactive Case Management (ICM) example to illustrate each approach. One approach is to find a mathematical optimum for multiple objectives, within cost constraints. In this approach, objectives are parameterized as functions, such as various response quality functions, and a cost function. Simulations can be a useful complementary analysis to validate the mathematical and statistical methods. In simulations using survey data, a set of features is modified and applied in the form of "what-if" scenarios. The trial and error approach addresses the main limitations of the mathematical and simulation-based optimization methods by varying features in the ASD, implementing these variations, and evaluating them for future implementation. One-time surveys are limited in what can be tried, evaluated, and used to further optimize the design. It is important to treat these approaches as complementary rather than alternatives. Each has unique strengths as well as limitations.
