The Search for Spatial Associations

The outline of the module is as follows:

1. Readings
2. Motivation: effective sample size
   1. One interpretation of spatial autocorrelation
   2. The univariate case
   3. The bivariate case
3. Decision-making in the presence of spatial autocorrelation: Spatial statistical theory simulation and matrix results
   1. Variance inflation/deflation factors
   2. Confidence intervals for means
   3. The significance of correlation coefficients
4. Graphical depictions
   1. The relationship between the SAR autoregressive parameter value and Moran Coefficient
   2. The relationship between the univariate effective sample size and the SAR autoregressive parameter estimate
   3. The relationship between the bivariate effective sample size and a pair of SAR autoregressive parameter estimates
5. Some useful equations
   1. Predicting the SAR autoregressive parameter value from a Moran Coefficient
   2. Predicting the univariate effective sample size from an SAR autoregressive parameter estimate
   3. Predicting the bivariate effective sample size from a pair of SAR autoregressive parameter estimates
6. Applications: Haining's Glasgow disease incidence data
7. Summary

Dr. Griffith's materials

• take our

• Try the lab!
• Additional Readings
• Please: evaluate this module. We appreciate your comments!
• Acknowledgement:
• Back to the
  

Page by Andy Long. Comments appreciated.