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: st: comparing coefficients using suest (same sample, same DV)
From
Richard Williams <[email protected]>
To
[email protected], [email protected]
Subject
Re: st: comparing coefficients using suest (same sample, same DV)
Date
Tue, 14 Aug 2012 10:00:53 -0500
At 06:30 AM 8/14/2012, Maarten Buis wrote:
On Tue, Aug 14, 2012 at 2:31 AM, Sauermann, Henry wrote:
> I estimate two regressions using the same sample:
>
> (1) Y=b1x1+e1
> (2) Y=b1x1+b2x2+e2
>
> and would like to test the difference in b1 across the two models.
This is what "SEM-people" call the indirect effect of x1 through x2 on
y. You can get that by estimating your model in Stata with -sem- and
than use -estat teffects- to see the decomposition of the total effect
of x1 into a direct and indirect effect. See -help sem- and -help
estat teffects-.
Hope this helps,
Maarten
Of course, it could be that x2 is a common cause of x1 and x2, in
which case, instead of it being an indirect effect, it would
represent spurious association. But for Henry's purposes he could
treat it as an indirect effect, and do something like this, I think:
use "http://www.indiana.edu/~jslsoc/stata/spex_data/ordwarm2.dta";, clear
sem (yr89 -> ed) (yr89 ed -> warm) , nocapslatent
estat teffects
Maarten, am I correct in saying that, if the indirect effect of the
x1 variable (in this case yr89) is statistically significant, you
would conclude that b1 is different in the two models? Or is there
some other aspect to this that I am missing?
-------------------------------------------
Richard Williams, Notre Dame Dept of Sociology
OFFICE: (574)631-6668, (574)631-6463
HOME: (574)289-5227
EMAIL: [email protected]
WWW: http://www.nd.edu/~rwilliam
*
* 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/