Hello,

My colleague and I are working on a latent class analysis of roughly 30 different behaviors. The LCA indicated we had four classes. In our next step, we would like to compare the average of each behavior between the four classes to expose significant differences after the
lcmean
test. The problem is that STATAs output provides a standard error for these coefficients. We know of the classic z-score equation of (b1-b2)/sqrt(SEb1-SEb2) but we would prefer not to do this by hand. After all, it would require over 180 individual computations. Is there any way that anyone knows of doing this quickly in STATA or another program for that matter? Or is there a way to get a standard deviation in the output for the use of a
ttesti
command?

Thanks!