ABSTRACT
Chapter 3 discusses the distance decay behavior, often referred to as the “first law of geography”, in patients seeking medical services. Such a behavior can be captured by various functions on how the volume of patient trips declines with increasing distance or travel time. Obviously, the measure of distance or travel time utilizes results from Chapter 2. This chapter illustrates (1) what functions to consider (e.g., power, exponential, square-root exponential, Gaussian, log-normal, log-logistic, and compound power-exponential), (2) how to identify the best-fitting one (e.g., by estimating a spatial interaction model or via fitting a complementary cumulative distribution function), and examines (3) what we can learn from the variability of the function across subpopulations and across geographic areas. Many spatial analysis tasks, including the Huff model in Chapter 5, rely on the result of derived distance decay function in modeling interactions between the supply and demand sides of a health care market.
