This course introduces spatial data analysis applied in multiple social science disciplines including sociology, demography, epidemiology, public health, political science, and biostatistics. The course consists of two modules: (1) exploratory spatial data analysis (ESDA); and (2) spatial modeling. Students will learn how to carry out ESDA and spatial modeling in different software packages including R, GeoDa, QGIS, and Stata.

In the ESDA module, we will introduce the three types of spatial data (areal data, point-referenced or geostatistical data, and mapped point process or point pattern data). We will learn basic process spatial data processing in preparation for analysis, Monte Carlo simulation and Bayesian spatial statistics, cluster analysis of mapped point process, and the concept of spatial autocorrelation.

In the spatial modeling module, we will survey various statistical models that suit different types of spatial data. Some examples are: (1) space-time models for mapped point process; (2) the conditional autoregressive model (CAR) and the simultaneous autoregressive model (SAR) for areal data; (3) Bayesian hierarchical models for point-referenced or geostatistical data; and (4) multilevel and Bayesian spatial models for small area estimation.

Throughout the course, we will focus on the intuition behind spatial statistics, the use of computer software, and coding. We will run computer simulations and replicate case studies in real social science research. Some of the case studies will be drawn from my own research.