- LISREL
- original and possibly still most popular program for SEM.
My personal favourite. Recent versions are much
easier to use (the new SIMPLIS language, and path diagram input are
both available - but
fear not, the old way of specifying matrices is still there.)
Later
versions have some MLM capabilities.
-
Mplus - a relative newcomer to the world of SEM, it has a
nice input language. Good for categorical and ordered
categorical analysis. Can also do MLM. I wrote a
review of version 2.0 in the British
Journal of Mathematical and Statistical Psychology
(volume 54, part 2, page 383-387; 2002).
- EQS - I have never been a
huge fan of EQS, but lots of people do
like it, so maybe that's just me. Also has some MLM
capability.
- AMOS - whilst I
always suggest that I like LISREL the most, it always seems to be AMOS
that I actually use. Nicest interface of all
the programs I know. Estimation with means is nice,
bootstrapping
is v. good. Was way ahead of the rest on missing data
analysis,
but the others have all just about caught up. I reviewed this
in
the
British
Journal of Mathematical and Statistical Psychology
(Volume: 53 Part: 1 Page: p141 -144 ).
- Mx
- you don't see Mx mentioned a great deal in lists of SEM
software, which I think is a shame. Partly because it's
rather a natty bit of software, with some nice features. And
partly because it's free. It's a bit tricky to use, until you
get the hang of
it, but once you have, it's not especially tricky.
If your
software isn't here, and you think it should be, see
this page
.
I was going to put a list of multilevel modelling / hierarchical linear
modelling software here, but there's a much better list here
.
|