Advanced GIS-A

Friday, May 23, 2014, 9:00am – 4:30pm
Presented by Jennifer Miller
Located in CLA 1.402
In track Vector

Spatial Statistical Analysis and GIS

This course extends the information covered in an introductory GIS course by exploring how GIS can be used to explore spatial relationships between real world phenomena and their relative locations. In addition to covering basic spatial statistical analysis, this course focuses on three fundamental aspects of statistical analysis: quantifying spatial patterns, identifying clusters, and analyzing geographic relationships. While these methods will be appropriate for many different types of data and inductive research questions, the in-class analysis will cover socio-economic and health applications.

Contents of the course:

I: Introduction to Spatial analysis and Point pattern analysis

1st order vs 2nd order effects (density vs. distance)
Nearest neighbor analysis, Ripley’s K statistic

  • Interpreting results (z-scores, Monte Carlo simulation)
  • Issues that affect results

II: (Global) Spatial pattern measurement and modeling

Spatial autocorrelation analysis

  • Conceptualizing spatial relationships
  • Moran’s I, Getis-Ord General G statistic
  • Interpreting results
  • Issues that affect results

OLS Regression

  • Model fitting
  • Model diagnostics
  • Residual analysis

III: (Local) Spatial pattern measurement and modeling

Local spatial autocorrelation analysis

  • LISA, Gi* (Geoda software)
  • Mapping/interpreting results
  • Issues

Geographically weighted regression

  • Spatial non-stationarity