|
Message/Author |
|
Cherry Chu posted on Thursday, November 02, 2017 - 8:13 am
|
|
|
Hi, I am conducting two-part GMM with covariates for 27 time points, and hope to compare growth curves between the experimental and control groups. To achieve this, is it okay to run models separately for each group, or must I combine them into one model using KNOWNCLASS? Thank you. |
|
|
You should combine them so that you can estimate differences in parameters across the groups. |
|
Back to top |
|
|