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: Re: comparing regression discontinuity treatment effects for different subsamples


From   Austin Nichols <austinnichols@gmail.com>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: Re: comparing regression discontinuity treatment effects for different subsamples
Date   Tue, 12 Oct 2010 22:41:12 -0400

Prashant <presearchwork@gmail.com>:
Yes, but what are fixed_bw1 and fixed_bw2 then?
And the Imbens and Kalyanaraman approach gives an optimal bandwidth of 2 here?

On Tue, Oct 12, 2010 at 9:14 PM, Prashant <presearchwork@gmail.com> wrote:
> Dear Austin,
>
> Thank you so much again. The cutoff in the assignment for males and
> females is the same. I was planning to use Imbens and Kalyanaraman
> (2009)'s automatic bandwidth procedure (as also kindly suggested by
> Juan) to first estimate bandwidths for male and female subpopulations
> separately and then insert those fixed bandwidths into the wrapper
> program. Is this okay?
>
> best,
> Prashant
>
> References:
> Guido Imbens,Karthik Kalyanaraman  (2009). Optimal Bandwidth Choice
> for the Regression Discontinuity Estimator. NBER Working Paper No.
> 14726.
>
> http://www.economics.harvard.edu/faculty/imbens/software_imbens
>
>
>
>
> On Wed, Oct 13, 2010 at 4:12 AM, Austin Nichols <austinnichols@gmail.com> wrote:
>> Prashant  <presearchwork@gmail.com>:
>> That looks like a good start, but what are you supplying to option z0?
>> Do males and females have different cutoffs in the assignment variable
>> (AKA running variable)?
>>
>> On Tue, Oct 12, 2010 at 9:19 AM, Prashant <presearchwork@gmail.com> wrote:
>>> Dear Austin and John,
>>>
>>> Thank you so much for your kind advice. I am new to programming in
>>> Stata and so would appreciate any further feedback on whether the
>>> following example code correctly captures the idea:
>>>
>>> capture program drop mycompare1
>>> program mycompare1, eclass
>>>        version 11.1
>>>        syntax varlist(numeric)
>>>        quietly rd `varlist' if female == 1, z0(fixed_bw1) mbw(100) bw(2) kernel(tri)
>>>        matrix treat1 = e(b)
>>>        quietly rd `varlist' if female == 0, z0(fixed_bw2) mbw(100) bw(2) kernel(tri)
>>>        matrix treat2 = e(b)
>>>        matrix diff = treat2 - treat1
>>>        ereturn post diff
>>> end
>>>
>>> bootstrap, reps(1000): mycompare1 yvar running_var
>>>
>>> Many thanks again--best wishes,
>>> Prashant

*
*   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