|
|
 |
| Multiple groups by sex, constraining |
 
|
| Message/Author |
|
|
|
|
Hello,
I am doing a LGC multiple groups analysis to look at different trajectories of depression over 6 time points. I will then add a time-varying predictor at 4 of the time points and I would like to see if the association between that time-varying predictor on depression is different for males vs females.
My questions are:
1) Do I start with a completely unconstrained model and then constrain certain pathways?
2) Would I start first with just the trajectory model, and then once that has a good model fit, then add in the time-varying predictor and see if those paths are different?
Thank you for your assistance.
Nicole |
|
|
|
|
| General analysis strategy questions are suitable for SEMNET. |
|
|
| Back to top |
|
|
