3

I'm somewhat confused about regression models. I decided to create a GWR model for a project using 35 polygons of health data. However I find now that my model has flaws and I cannot create a GWR model probably because the dataset is too small.

When creating a OLS are we meant to remove all variables that are related? if I'm modelling obesity, I would think that income, exercise are factors that relate to it so what purpose does the OLS model serve? I created a model after removing auto correlated variables but I'm kind of confused as to what its meant to be showing.

Sy_n
  • 31
  • 1
  • 2
  • 1
    This might help answer your question.

    http://gis.stackexchange.com/questions/27245/regression-model-and-spatial-autocorrelation

     – Sam007 Jun 13 '12 at 00:15

1 Answers1

1

This might help answer your question.

Regression Model and Spatial Autocorrelation

It gives you the understanding of how the OLS and GWR are statistically different.

Community
  • 1
Sam007
  • 4,395
  • 7
  • 48
  • 69