[R-sig-ME] Comparing the effect size of parameters across non-nested models with different outcomes

[Ben Bolker]

Isaac Petersen <dadrivr at ...> writes:

> 
> Hi,
> 
> Is there a way to compare the effect size of parameters across non-nested
> models with different outcomes?  I would like to test statistically whether
> one parameter is larger than another.  I'm currently using nlme, so that
> would be preferable if possible.  Thanks guys!


 You could standardize both your predictors and your outcomes to all
have standard deviations of 1, then do a t-test between the effects ...