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 on April 23, and its replacement, statalist.org is already up and running.


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

Re: st: SV: RDD robustness check: Controlling for covariates?


From   Austin Nichols <austinnichols@gmail.com>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: SV: RDD robustness check: Controlling for covariates?
Date   Tue, 14 Feb 2012 09:31:48 -0500

Miriam Wüst <miw@sfi.dk>:
But the linear effect of the covariate can vary on either side; the
assumption is that the mean level of the covariate does not jump at
the discontinuity, not that its coef does not. In fact, the treatment
regime could affect the effects of other variables. I would allow a
covariate to have different effects on both sides of the
discontinuity, but I would be very careful to ensure to redefine the
covariate so that the mean level of the covariate is zero at the
discontinuity, so that no effects of the covariate are loaded onto the
dummy measuring the jump in mean outcomes.  This is a dangerous
exercise, though, since you are admitting more potential bias in a
quest for greater efficiency, and you could actually wind up with
lower efficiency...

On Tue, Feb 14, 2012 at 2:44 AM, Miriam Wüst <miw@sfi.dk> wrote:
> It would not be great if the effect of your covariate varies on the two sides as you assume that there is no discontinuity in other variables at the theshold...
>
>
> -----Oprindelig meddelelse-----
> Fra: owner-statalist@hsphsun2.harvard.edu [mailto:owner-statalist@hsphsun2.harvard.edu] På vegne af Jen Zhen
> Sendt: 13. februar 2012 17:24
> Til: statalist@hsphsun2.harvard.edu
> Emne: st: RDD robustness check: Controlling for covariates?
>
> Dear list members,
>
> just as a robustness check (not for the main estimates) for a regression discontinuity design, I'd like to test whether results change when I control for a covariate.
>
> However, I'm unsure whether I should just control for the covariate in the same way on both sides of the threshold, or whether I need to allow the effect of the covariate to vary between the two sides, i.e.
> should I use
>
> - reg outcome T forcingvar T_forcingvar covar - or
> - - reg outcome T forcingvar T_forcingvar covar T_covar-  ?
>
> Thanks so much and best regards,
> JZ

*
*   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