C.T.Shehzad wrote:
>Actually stata has routine for cross sectional factor analysis (its
command in stata is - factor - ).>
OK, I know what you mean now. Yes indeed, -factor- is decent but missing
some things I would want (e.g., MINRES/OLS fitting, standard errors for
the loadings).
>>But I think it does not have capability for panel data.
Not really, no, unless you go back to very "old skool" fitting
procedures that combine Procrustes rotations with ordinary extraction
methods.
>>So I was just asking if someone has tried to
program factor analysis for panel data. Factor analysis or principal
component analysis is a technique for data reduction.<<
This stuff goes under the general name "multiway factor analysis," at
least in psychometrics. I don't know of Stata programs for it, but may
be wrong (Googling turned up a concept paper and that was it). If you're
talking about a confirmatory model you might be able to set that up in
-gllamm-.
JV
*
* 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/
