Including mean structure
Hi
Thanks for the recent clarification about invariance testing. I have a follow-up question. I tried to recreate the Holzinger-Swineford example from the tutorial. All is good except for one thing. For some odd reason, clicking on "include mean structure" does not seem to actually add latent the means. My chi-square values for the first two models (configural and metric) invariance match those in the tutorial, but the df are different (as are the AIC and BIC values). Also, the means for the three latent variables is zero (in all three models). Fixing the intercepts in the third (scalar invariance) model works, but the df still don't match those in the tutorial. I'm likely just missing something obvious, but I wanted to be sure there wasn't a bug with the 'include mean structure" on the latest JASP version.
Thanks again for your help!
Seth
Comments
Hi
Sorry. . . I figured this out. I was trying to compare groups on the latent mean after doing invariance testing. I added the necessary lavaan code to test both in in JASP.
Seth
:-) great