LTA with 2 covariates and interaction...
Message/Author
 Anya Hart posted on Sunday, September 08, 2019 - 1:49 am
I am wondering what is the best way to have 2 covariates and the interaction between them in an LTA.

Has anyone tried this?

Right now, I am thinking about transforming one of my covariates into a binary variable so that I can use the KNOWNCLASS option on top of example 8.14 like this:

CLASSES = cg (2) c1 (3) c2 (3);
KNOWNCLASS = cg (g = 0 g = 1);

MODEL:

%OVERALL%
c1 c2 ON cg;

MODEL cg:

%cg#1%
c2 ON c1;
c1 ON x;

%cg#2%
c2 ON c1;
c1 ON x;

MODEL c1:

%c1#1%
c2 ON x;
[u11\$1] (1);
[u12\$1] (2);
[u13\$1] (3);
[u14\$1] (4);
[u15\$1] (5);

%c1#2%
c2 ON x;
[u11\$1] (6);
[u12\$1] (7);
~
[u15\$1] (15);

MODEL c2:

%c2#1%
[u21\$1] (1);
[u22\$1] (2);
[u23\$1] (3);
[u24\$1] (4);
[u25\$1] (5);

%c2#2%
[u21\$1] (6);
[u22\$1] (7);
[u23\$1] (8);
~
[u25\$1] (15);
 Bengt O. Muthen posted on Sunday, September 08, 2019 - 4:56 pm
Why not simply use

Define: x12 = x1*x2;

and use x12 instead of x in your setup (so no cg).