0
$\begingroup$
mtcars=data.frame(mtcars)
m=glm(vs~mpg+cyl+disp+hp+drat,family="binomial",data=mtcars)
m=glm(vs~wt,family="binomial",data=mtcars)

I seek to estimate pseudo-r-squared that is adjusted for number of predict values however I have not seen such a resource.

Is there a way to do so?

$\endgroup$
1
$\begingroup$

You could manually calculate it as 1-(Residual Deviance/Null Deviance). Both deviances are in your model output.

| cite | improve this answer | |
$\endgroup$
  • $\begingroup$ this does not count for one model have much more covariates than the other $\endgroup$ – bvowe Apr 6 '19 at 13:11

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