Sorry! I made a mistake with my example. I meant to say group_effect instead of rand_effect. It should read as follows:
Y1 = B1 + B2(time) + B3(group_effect1) + B4(group_effect*time) + B5(covariate_1) +B6(covariate_2) Y2 = B1 + B2(time) + B3(group_effect1) + B4(group_effect*time) + B5(covariate_1)
The difference between the two that I was trying to highlight is the presence of the second covariate term.
Thanks, -E
On Tue, Mar 4, 2014 at 11:18 AM, Eric Cunningham etc42@hawaii.edu wrote:
Hello Freesurfer experts, I have a question about how to determine if a covariate is important in a mixed effects model for example:
Y1 = B1 + B2(time) + B3(rand_effect1) + B4(rand_effect*time) + B5(covariate_1) +B6(covariate_2) Y2 = B1 + B2(time) + B3(rand_effect1) + B4(rand_effect*time) + B5(covariate_1)
Would it be appropriate to use the lme_mass_LR script? (the script says it is for comparing models with q and q+1 random effects).
Thanks, -E