below
Gary
--
*Gary King* - Albert J. Weatherhead III University Professor - Director,
IQSS - Harvard University
GKing.Harvard.edu <http://gking.harvard.edu/> - King(a)Harvard.edu -
@kinggary<http://twitter.com/kinggary>- 617-500-7570 - Asst 495-9271 -
Fax 812-8581
On Sat, Apr 23, 2011 at 11:39 PM, Leslie Finger <lfinger(a)fas.harvard.edu>wrote;wrote:
Hi Class,
Sorry for all the email harassment. I am using fixed effects for villages
with
an IV. I have a few questions:
1) This doesn't work with ivreg, so I'm just doing each stage separately
(i.e.
taking the fitted values of the first stage and putting them into the
second
stage with all the same controls.) Is that valid? Am I cheating?
you should get the same coefficients, but the standard errors will be wrong
because you're telling the computer program that you know (rather than are
estimating) the predicted value from the first stage.
2) While it does report the coefficients, I'm getting messages like this "2
not
defined because of singularities". I checked, and these are some of the
villages with very few numbers of observations. Should I just ignore this
and
proceed?
don't ignore it. its telling you you have nonidentification, probably due
to collinearity. i'd respecify in some way.
Thanks!
-Leslie
_______________________________________________
gov2001-l mailing list
gov2001-l(a)lists.fas.harvard.edu
http://lists.fas.harvard.edu/mailman/listinfo/gov2001-l
_______________________________________________
gov2001-l mailing list
gov2001-l(a)lists.fas.harvard.edu
http://lists.fas.harvard.edu/mailman/listinfo/gov2001-l