Dear Mike,
I have a question about possibilities for model specification at stage 2 of MetaSEM.
So I have a dataset with 10 variables - at the end of stage 1 I get a pooled correlation matrix. In stage 2 I would like to test some models with different constellations of the variables, e.g. with variables V1, V2, V3, V4, V5, V6, V7 and V8 in the first model and with V1, V2, V3, V4, V5, V9, V10 in the second. Is that possible? And if it is possible, how should I specify this in the syntax?
Thanks for your help!
Printer-friendly version
Yes. But you have to select the variables you needed in the stage 2 analysis. Attached is an example.