Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: FW: st: SEM


From   "JVerkuilen (Gmail)" <[email protected]>
To   [email protected]
Subject   Re: FW: st: SEM
Date   Thu, 11 Oct 2012 22:12:54 -0400

On Thu, Oct 11, 2012 at 9:10 PM, Tucker, Graeme (Health)
<[email protected]> wrote:
>
> Brilliant. Thank you.

This model is identified and has exactly the same final log-likelihood
as the one you specified.

. sem (L1 -> a1 a2 a3 a4 a5 c1 c2 c3 c4 c5) (L2 -> a2 a3 a4 a5 c1 c2
c3 c4 c5) , covstruct(_lexogenous, diagonal) latent(L1 L2) stand

I constrained the loading for a1 on L2 to be 0. McDonald (1999) calls
this "echelon form" and it allows you to identify an orthogonal EFA
solution to get estimates using a CFA program. You can save the
resulting loadings and rotate using -rotatemat-. An interesting
alternative that is also described in McDonald (1999) is
"semi-confirmatory" factor analysis. In this case you identify each
factor with some factorially pure items and then allow all the rest to
load on all factors. For instance, if there were five other items in
this dataset you were uncertain about, load all the a items on L1, the
c items on L2 and the remaining items on L1 and L2.

McDonald, R. P. (1999). Test Theory: A Unified Treatment. LEA.
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/


© Copyright 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index