Statalist The Stata Listserver


[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

Re: st: Propensity Score Matching


From   "Cristian Mackenna" <chicomackenna@hotmail.com>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: Propensity Score Matching
Date   Wed, 18 Oct 2006 15:02:14 +0000

The different commands only show the full sample and not the matched ones.

In fact attk, atts, etc show always the full sample (example 30.000) even if the covariant vars have missing values and the probit obs are (example 25.000)

Psmatch2 always shows the complete observations of the probit (25.000 in this case) even if you are running nearest neighbor, epan, etc…

Running Psmatch2

psmatch2 dsip, outcome(com) pscore(pscore) kernel kerneltype(normal) ate

----------------------------------------------------------------------------------------
Variable Sample | Treated Controls Difference S.E. T-stat
----------------------------+-----------------------------------------------------------
com Unmatched | 295.504222 245.775165 49.7290569 1.38890102 35.80
ATT | 295.504222 256.261943 39.2422789 1.80287861 21.77
ATU | 245.775165 288.972377 43.1972121
ATE | 42.9481965
----------------------------+-----------------------------------------------------------
Note: S.E. for ATT does not take into account that the propensity score is estimated.

| psmatch2:
psmatch2: | Common
Treatment | support
assignment | On suppor | Total
-----------+-----------+----------
Untreated | 21,996 | 21,996
Treated | 1,478 | 1,478
-----------+-----------+----------
Total | 23,474 | 23,474

Runnig att

attk com dsip, pscore(pscore)

ATT estimation with the Kernel Matching method

---------------------------------------------------------
n. treat. n. contr. ATT Std. Err. t
---------------------------------------------------------

1852 32563 39.242 . .

---------------------------------------------------------
Note: Analytical standard errors cannot be computed. Use
the bootstrap option to get bootstrapped standard errors.




From: "jean ries" <jean.ries.l@gmail.com>
Reply-To: statalist@hsphsun2.harvard.edu
To: statalist@hsphsun2.harvard.edu
Subject: Re: st: Propensity Score Matching
Date: Wed, 18 Oct 2006 10:25:43 +0200

Difficult to guess what's wrong. Maybe you should show use what you
typed and the relevant portions of the output.

Here is some further advice:

* Make sure that the the commands are equivalent, i.e. expected to do
the same thing. Read carefully through the psmatch2 help-file.

* Make sure that you have not selected a "common support" option in
one command and not in the other.

* When there are if or in clauses in the commands, make sure that they
select the same observations.

* Make sure that everything is up-to-date.

Cheers,

Jean

On 10/18/06, Cristian Mackenna <chicomackenna@hotmail.com> wrote:
Did someone now why with the same sample and the same pscore varname, if I
run

psmatch2 with Gaussian kernel

the untreated group and the treated group have deferent's number that if I
run

attk

even if the ATT calculated is the same

*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/
_________________________________________________________________
Express yourself instantly with MSN Messenger! Download today it's FREE! http://messenger.msn.click-url.com/go/onm00200471ave/direct/01/

*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* 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   |   What's new   |   Site index