Event Title | Applied Spatial Regression Analysis |
Location | Davis Rm. 219 |
Sponsor | H.W. Odum Institute |
Date/Time | 03/27/2017 - 03/29/2017 1:30 PM - 4:00 PM | Event Price |
This short course provides an introduction to the field of spatial regression modeling. When analyzing data aggregated to geographic areas (e.g., census data for counties), a fresh set of issues arise that are not present in traditional non-spatial data analyses. These issues need to be recognized and accounted for when properly specifying regression models using attributes that are linked to geographic location. The topics covered in two afternoon sessions include:
• Why standard regression models generally fail when analyzing spatial data
• Defining and understanding “spatial autocorrelation”
• Causes of spatial autocorrelation
• Measuring & operationalizing spatial effects
• Defining spatial “neighborhoods”
• Creating spatial weights matrices
• Moran’s I statistic
• Incorporating spatial effects in spatial regression models
• Specification & estimation of spatial regression models
• Spatial regression model diagnostics
• (Time permitting: some interesting extensions to related topics)
Examples of estimating spatial regression models will use the open source software suite R (no prior knowledge of R is necessary)
Registration Fees: