Re: st: Stata 12 Announcement

["Joseph Coveney" <jcoveney@bigplanet.com>]

Tony Lachenbruch wrote:

Amen to the MPlus commentary.  Awful user interface, mysterious output, and
tests of variances using normal distribution.  What were they thinking?

Clyde Schechter wrote:

I think Stata Corp. has really hit a home run with this release.  I've
been around since version 4, and this looks like the most new capabilities
(and, from my personal perspective, the most useful new capabilities) in
that entire era.

I'm particularly looking forward to not having to use M-Plus any
more--really dislike that software.

--------------------------------------------------------------------------------

I'm very impressed, too, and I join in congratulating StataCorp for -sem- and
its GUI.  They're one heck of an accomplishment, and I look forward to using
them.

But I won't go so far as to knock Mplus just yet.

Yes, it has a BMD-ish command language, and, yes, it's difficult or impossible
to get at some of the results electronically.  

But it can fit SEMs with categorical indicator variables by maximum likelihood,
that is, without resorting to ADF/GMM or to treating them as normal.  And it can
handle models with latent classes.  It's not sitting still, either, with
recently introduced "exploratory SEM" and some Bayesian capabilities.

It still has a place in the armamentarium.

Joseph Coveney


*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/