I have a question concerning the use of latent variables at both the within and between level using multilevel SEM.
In my model statement, I define my latent variables using the following input:
%within% TK BY TK1 TK2 TK3 TK4 TK5 TK6; PL BY PL2 PL3 PL4 PL5; MT BY MT1 MT2 MT3 MT4 MT5 MT6 MT7; ZEPROD BY ZEPROD1 ZEPROD2 ZEPROD3; ZEPROC BY ZEPROC1 ZEPROC2 ZEPROC3 ZEPROC4;
This works perfectly, but I need the same latent variables again at the between level. I want MPlus to use the latent structure as defined at the within level, as I want it to be based on individual scores rather than group means.
I tried it with the following ouput at the between level, but this is not working:
%between% TKB BY TK1 TK2 TK3 TK4 TK5 TK6; PLB BY PL2 PL3 PL4 PL5; MTB BY MT1 MT2 MT3 MT4 MT5 MT6 MT7; ZEPRODB BY ZEPROD1 ZEPROD2 ZEPROD3; ZEPROCB BY ZEPROC1 ZEPROC2 ZEPROC3 ZEPROC4; SE_CM BY SE_CM1 SE_CM2 SE_CM3 SE_CM4;
Could you please help me in figuring out how I should resolve this?