Hi all,
I am estimating a multilevel multinomial logistic model with GSEM and do not fully understand constraining random effects and how to interpret these effects.
Here is a simple null model: gsem (i.speed_pattern <- Le1[tid]), mlogit
The result is:
My questions are:
1) what is the meaning of constraining random effect for 1.speed_pattern, and how do I understand it?
2) what is Le1[tid] (random effect) in other classes? Is it the average difference between group intercepts and the overall intercept? Or standard deviation, or other things?
3) I guess var(Le1[tid]) considers five classes. Does it calculate from Le1[tid] at different classes?
Thank you in advance!
Best,
Hong
I am estimating a multilevel multinomial logistic model with GSEM and do not fully understand constraining random effects and how to interpret these effects.
Here is a simple null model: gsem (i.speed_pattern <- Le1[tid]), mlogit
The result is:
My questions are:
1) what is the meaning of constraining random effect for 1.speed_pattern, and how do I understand it?
2) what is Le1[tid] (random effect) in other classes? Is it the average difference between group intercepts and the overall intercept? Or standard deviation, or other things?
3) I guess var(Le1[tid]) considers five classes. Does it calculate from Le1[tid] at different classes?
Thank you in advance!
Best,
Hong
