Chapter 19 LISA for Discrete Variables

So far, the application of local spatial autocorrelation statistics has been to continuous variables. In this Chapter, discrete variables are considered, and, more specifically, binary variables. To address this context, a univariate Local Join Count statistic (Anselin and Li 2019) and its extension to a multivariate setting are introduced. The latter allows for a distinction between situations where the two discrete variables can co-occur (i.e., take the value of 1 for the same location), and where they cannot (no co-location).

The principle behind the Local Join Count statistic is broadened by applying it to a subset of observations on a continuous variable that satisfy a given constraint. The most common application of this idea is to a specific quantile of the observations, leading to the concept of a Quantile LISA (Anselin 2019b).

The Chicago SDOH sample data set is again used to illustrate these methods.