|
![](http://www.statmodel.com/discussion/icons/csc/1px.gif) |
Comparing R-squaured values |
![Previous](http://www.statmodel.com/discussion/icons/csc/navgen_arrow_left.gif) ![](http://www.statmodel.com/discussion/icons/csc/1px.gif)
|
Message/Author |
|
|
I have tested two models using the TYPE=COMPLEX procedure, where the first model contains only covariates and a second model contains the covariates plus additional substantive predictors. I am trying to mimic a hierarchical regression where one would calculate an R-squared change. Is it reasonable to assume that we can statistically compare the R-square values for each model like one would with a partial F-test in multiple regression or are there other considerations when using the TYPE=COMPLEX procedure? |
|
|
I don't know how, but it seems like you should be able to do an analogous procedure. |
|
Back to top |
|
|