Bookmark and Share

Notice: On March 31, it was announced that Statalist is moving from an email list to a forum. The old list will shut down at the end of May, and its replacement, statalist.org is already up and running.


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

Re: st: Use of polychoricpca with imputed data


From   Francesca Pesola <PesolaF@cardiff.ac.uk>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: Use of polychoricpca with imputed data
Date   Thu, 8 Mar 2012 18:04:37 +0000

Thanks for your reply.

I used the -category(manip)- option and it seems to have worked; what I am not sure about is how well. 

I am working with survey data and have a number of alcohol items which I need to combine; unfortunately, they are not from a standard questionnaire which is why i set off to use -polychoricpca-. 

Once I run the PCA, I use the factor scores as a predictor in a number  of regressions.  When I compare the regression output for the imputed data to that of the complete cases, results are similar. 

Do you think for the purpose of my analyses, the use of -category(manip)- is appropriate?

Could you recommend some articles and/or sites on how to use the -program, eclass- command?

Thanks again.
-----owner-statalist@hsphsun2.harvard.edu wrote: -----
To: statalist@hsphsun2.harvard.edu
From: Stas Kolenikov 
Sent by: owner-statalist@hsphsun2.harvard.edu
Date: 03/08/2012 05:34PM
Subject: Re: st: Use of polychoricpca with imputed data

It is not going to work out satisfactorily. -polychoric- is not an
estimation command, in the sense of producing proper -e(b)- and -e(V)-
matrices that -mim- could pick up. A reasonable way to proceed would
be to write your own -program, eclass- that would accept [pw] and [if]
conditions (that -mim- uses), would run -polychoricpca- on the sample
it receives, and use it in a further estimation command such as
-regress- (which would still produce underestimated standard errors
that won't account for sampling variability in the predicted scores).
If you want to get some sort of averaged scores, you can try
-category(manip)-, although I am not entirely sure what the
interpretation of the results would be. If you would explain the
analytic purpose of your exercise, may be I'd come up with better
suggestions.

I've been meaning to rewrite -polychoric- as a wrapper for -cmp-, so
that it would produce appropriate standard errors (which ain't rocket
science given how much work David Roodman put into his wonderful
package). Whether I will ever do that is an open question.

On Thu, Mar 8, 2012 at 5:31 AM, Francesca Pesola <PesolaF@cardiff.ac.uk> wrote:
> I have imputed my dataset using the 'mi ice' command in Stata 11 and I would like  to run a principal component analysis, using the 'polychoricpca' command, on the imputed data.  I am interested in saving the extracted component to the dataset to use it in my analyses.
>
> I used mim2: polychoricpca varlist, pw score(outcome) nscore(3)
>
>
> But I received the following message:
>
> command polychoricpca not recognised by mim; try specifying category option
> r(198);
>
>
> How can I specify the category option?
> Cheers,
> Francesca
>
> *
> *   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/



-- 
Stas Kolenikov, also found at http://stas.kolenikov.name
Small print: I use this email account for mailing lists only.

*
*   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/

*
*   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/


© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   Site index