[R-sig-ME] R-sig-mixed-models Digest, Vol 82, Issue 25

[Lenth, Russell V]

If I understand this correctly, I think you can use the 'trend' argument in lsmeans to get what you want. See the man page for lsmeans and the examples near the end.

Russ Lenth

Sent from my iPad

> On Oct 16, 2013, at 2:27 PM, "r-sig-mixed-models-request at r-project.org" <r-sig-mixed-models-request at r-project.org> wrote:
> 
> Thanks guys. I am using lsmeans, however the problem is that I am  
> interested in a categorical:continuous predictor interaction so I have  
> two problems
> 
> 1. lsmeans seems to take one value of the continuous predictor and  
> compare the categorical predictors in relation to the outcome variable  
> on this one value, as opposed to comparing slopes of the variable; I  
> therefore wanted to graph this interaction somehow with the adjusted  
> means