FullModel<- (lm(Fubar~.-Foo-Bar,data=BarFoo)) NullModel<-(lm(Fubar~1)) step(NullModel,scope=formula(FullModel),direction="forward",k=log(nrow(BarFoo)))

When doing the above forward stepwise regression, the forward steps halt before certain variables are added in. does this mean that they do not improve the AIC score or does it mean that the inputs are in error?


1 Answer 1


That means that they do not improve AIC.

But forward selection isn't great (no automatic method is).


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.