|
|
Panel model with latent and manifest ... |
|
Message/Author |
|
|
Hi, I am currently trying to analyse a panel model in which there are both latent and manifest variables regressed onto one another at multiple time-points. For example, depression at time 2 as a latent variable is regressed onto obesity at time 1 as a manifest variable, and similarly obesity at time 2 is regressed onto depression at time 1. When I am investigating the measurement invariance of of the latent factors, should I be correlating the latent variables with the manifest variables using WITH statements, or use ON statements to regress onto the manifest variables as in example 5.8 from the user guide? |
|
|
I would use WITH for manifest and latent at the same time point. You may also consider RI-CLPM. |
|
|
Sorry for the slow response, and thanks for your reply. So to be clear, in the example above I would co-vary depression and obesity only at the same time-point and then have regression paths for the associations over time? |
|
|
Right. |
|
Back to top |
|
|