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: disattenuated regression


From   John Antonakis <John.Antonakis@unil.ch>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: disattenuated regression
Date   Sun, 18 Nov 2012 17:42:12 +0100

Hi:

MPlus is a tad (well to be honest a mile) ahead when it comes to SEM models having categorical indicators. Of course, this Stata's sem version 1, so we hope to see some developments in Stata in the line of:

Flora, D. B., & Curran, P. J. (2004). An empirical evaluation of alternative methods of estimation for confirmatory factor analysis with ordinal data. Psychological Methods, 9(4), 466-491. Muthén, B. O. (1984). A general structural equation model with dichotomous, ordered categorical, and continuous latent variable indicators. Psychometrika, 49(1), 115-132. Muthén, B. O., du Toit, S. H. C., & Spisic, D. (in press). Robust inference using weighted least squares and quadratic estimating equations in latent variable modeling with categorical and continuous outcomes. Psychometrika.

So, if you have such models it is best to use MPlus and the robust WLS (i.e., WLSMV) estimator.

As for your question, one can of course regress a latent variable on an observed variable. But it all depends on how you have done it (i.e., your full specification), by ensuring that the model is identified.

Best,
J.

__________________________________________

John Antonakis
Professor of Organizational Behavior
Director, Ph.D. program in Management

Faculty of Business and Economics
University of Lausanne
Internef #618
CH-1015 Lausanne-Dorigny
Switzerland
Tel ++41 (0)21 692-3438
Fax ++41 (0)21 692-3305
http://www.hec.unil.ch/people/jantonakis

Associate Editor
The Leadership Quarterly
__________________________________________


On 18.11.2012 17:33, 'Alim Beveridge wrote:
> Hi John,
> thanks again for your advice. I was thinking of doing this actually and have been teaching myself how to use sem in stata. > However does it deal with dichotomous/categorical IVs, like Mplus does? It appears that it only does ML and ADF, and my sample size is too small for ADF (N = 210). Someone on statlist described how to create your own correlation matrix and then run sem on that. is that how I would have to do it?
>
> also when I put a regression path between an observed IV (say age or hours worked) and my latent DV, the model does not converge. These observed vars are only in the structural part of my syntax, not the measurement. Is this not allowed?
> thanks,
> 'Alim
>
> On 11/17/2012 5:18, John Antonakis wrote:
>> That's easy. Use -sem-
>>
>> It supports vce(robust) and vce(cluster id)
>>
>> See -help sem-. Note, if you have a single indicator of a latent the basic set up would be (for predicting y from proxy x of x*, the latter being the latent):
>>
>> (note latents are denoted in capitals in sem unless you specify otherwise)
>>
>> sem (X->x@1) (y<-X) , reliability(x .80) vce(cluster id)
>>
>> HTH,
>> John.
>>
>> __________________________________________
>>
>> John Antonakis
>> Professor of Organizational Behavior
>> Director, Ph.D. program in Management
>>
>> Faculty of Business and Economics
>> University of Lausanne
>> Internef #618
>> CH-1015 Lausanne-Dorigny
>> Switzerland
>> Tel ++41 (0)21 692-3438
>> Fax ++41 (0)21 692-3305
>> http://www.hec.unil.ch/people/jantonakis
>>
>> Associate Editor
>> The Leadership Quarterly
>> __________________________________________
>>
>>
>> On 16.11.2012 21:13, 'Alim Beveridge wrote:
>> > eivreg does what I want but it does not work with the vce option. it does not support svy either. Does anyone know how to do regression in Stata with correction for measurement error (like eivreg) AND cluster robust SEs (like vce (cluster)).
>> >
>> > Thanks,
>> > 'Alim
>> >
>> > On 11/15/2012 4:01, Maarten Buis wrote:
>> >> On Thu, Nov 15, 2012 at 4:48 AM, 'Alim Beveridge wrote:
>> >>
>> >>> I would like to run disattenuated regression (DR) in Stata. I have not found
>> >>> a package that will do this for me.
>> >>>
>> >> Did you look at -help eivreg-?
>> >>
>> >> Hope this helps,
>> >> Maarten
>> >>
>> >> ---------------------------------
>> >> Maarten L. Buis
>> >> WZB
>> >> Reichpietschufer 50
>> >> 10785 Berlin
>> >> Germany
>> >>
>> >> http://www.maartenbuis.nl
>> >> ---------------------------------
>> >> *
>> >> *   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/
>> >>
>> >>
>> > *
>> > *   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/
>> *
>> *   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/
>>
> *
> *   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/
*
*   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 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   Site index