Spatial Regression Modeling

In 2011 GISPopSci held a five-day workshop on Spatial Regression Modeling. Led by Paul R. Voss (University of North Carolina at Chapel Hill) and Katherine Curtis (University of Wisconsin-Madison), the purpose of this workshop was to provide an overview of applied spatial regression analysis (or spatial econometrics). This course introduced the broader field of spatial data analysis as well as a range of issues applicable to analyzing georeferenced data.

Course Overview:

This workshop set out to address the following questions:

  • How does spatial autocorrelation arise?
  • How is it measured and understood?
  • How does it relate to issues of spatial heterogeneity and spatial dependence?
  • How should it inform the specification and estimation of regression models?

The course was structured around a lecture format combined with computing lab exercises. Although we used mapping software, the focus of the course was on spatial analysis, not geographic information systems (GIS). Software emphasis was given to GeoDa and R for exploratory spatial data analysis (ESDA) and spatial regression modeling. Some acquaintance with this software is helpful but is not a prerequisite. Prerequisites for maximizing learning in this course are a solid grounding in standard multivariate regression techniques and a minimal level of comfort with matrix notation and algebra.

Accessing the Course:

This content has been categorized into five modules. Each module contains readings and presentations or laboratory exercises. We have preserved this course in a number of different formats, including video recordings, audio recordings, lecture slides, and notes on the lectures. Each module is broken into multiple parts, and can be accessed through the links below. We recommend experiencing these segments by viewing the video files while paging through the lecture slides simultaneously. A faster way to experience this content is to review the presentation recap, while paging through the presentation