Re: [R] [R-sig-ME] multcomp package

Hi, It seems you want to test the fixed effect. So I think the related p-value is appropriate. I'm not sure... - Original Message - From: li li To: r-sig-mixed-models , r-help , epalm...@cns.umass.edu Subject: Re: [R-sig-ME] multcomp package Date: 20

Re: [R] [R-sig-ME] multcomp package

Thanks Evan and Gabriel for the reply. I think it might help me make the question clearer if I show the data and the model here (I actually asked questions related to this data before but I still need some help). The data looks like the following: response individual time method 1102.9

Re: [R] [R-sig-ME] multcomp package

Hanna, If you are interested in whether effect of time varies across individuals, why don't you fit a model with the predictors (time*individual) as fixed effects rather than random? On Mon, Jun 6, 2016 at 10:57 AM, li li wrote: > Hi all, > After fitting a random slope and random intercept mod