I have run GWR for a layer with polygons. For some features the Std Residual has high value as well as the local R-square.

How is this possible?

If std residual is high, the predicted value will be higher than the real, is it right?

So, the local R-square shouldn't be small as the regression didn't predict right the real value but the local R-square is high.

What is the misunderstanding?

  • I am not sure about the correct answer, maybe @whuber might help, R-square is a goodness of fit measure for the mean value against the modelled mean value and the actual values might be relatively more different than this mean, therefore larger standardised residuals. I think this is similar to conceptual difference between standard error and standard deviation. – fatih_dur Sep 20 '16 at 0:13

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.