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: Moderated mediation analyses


From   John Antonakis <John.Antonakis@unil.ch>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: Moderated mediation analyses
Date   Tue, 04 Dec 2012 08:47:49 +0100

Hi Isok:

Irrespective of your error, and similar to what I suggested to you previously, you are still using sureg, which will give you OLS estimates. If your mediator is endogenous and depends on the exogenous variable/s, it may correlate with the disturbance of your y equation. Thus, there is a possibility that your estimator is inconsistent. You need to do a Hausman test to determine whether the mediator correlates with the disturbance of y; ideally, if you are overidentified you should do a Hansen-Sargan test to ensure that the excluded instruments do not correlate with the disturbance of y. So, first estimate the basic model and see if it is correctly specified before trying to determine the conditional effects......though looking at your model, it seems that it is not identified (i.e., you need to exclude one instrument from the y equation in order to predict m).

Best,
J.

__________________________________________

Prof. John Antonakis
Faculty of Business and Economics
Department of Organizational Behavior
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 04.12.2012 06:03, William Buchanan wrote:
Check your syntax again. You're incorrectly referencing the estimate you're interested in. [y]_b[wx1] is what you might be trying to reference, but that estimate does not exist in the first equation [m1].

HTH

Sent from my iPhone

On Dec 3, 2012, at 19:14, "Kim, Isok" <isokkim@buffalo.edu> wrote:

Hello,

I am trying to replicate moderated mediation method (Preacher, Rucker & Hayes, 2007) using the UCLA Stata FAQ, (http://www.ats.ucla.edu/stat/stata/faq/modmed.htm#model5).  Similar to my previous inquiry about writing stata codes for mediation analysis, it involves -sureg- and -nlcom-. This time, I'm having trouble calculating conditional indirect effect.  Below is my stata code:

//RENAME vars for simplification
rename CESD y /* dependent variable */
rename PRD x1  /* independent variable */
rename GC3E m1 /* mediator variable 1 */
rename ES w /* moderator variable */

/*-sureg-&-nlcom- methods*/
set more off
summarize w
global m=r(mean)
global s=r(sd)
capture generate wx1=w*x1 /* moderator(ES) by IV(PRD) */
capture generate m1w=m1*w /* Mediator 1(GC3E) by Mod(ES) */

set more off
sureg (m1 x1 w i.SEX AGE EDU i.UNEMP i.MARSTS i.NATIVITY) ///
      (y x1 w wx1 m1 m1w i.SEX AGE EDU i.UNEMP i.MARSTS i.NATIVITY) if NMV5==1

//Conditional indirect effect via m1 with w
nlcom ([m1]_b[x1]+($m-$s)*[m1]_b[wx1])*([y]_b[m1]+($m-$s)*[y]_b[m1w])  /*mean - 1sd*/

Everything appears to run without an error until the last -nlcom- command. The error message reads, "[wx1] not found."  r(111).  Any suggestion or insight into resolving this would again be greatly appreciated!

Best,
isok

*
*   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