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: Propensity score matching: confidence intervals


From   Richard Goldstein <richgold@ix.netcom.com>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: Propensity score matching: confidence intervals
Date   Tue, 26 Jul 2011 15:53:49 -0400

yes, there is a robust option for logit -- it is a "vcetype" within the
vce option -- see -h logit- where you will see "vce(vcetype)" as an
option and you can click on vcetype to see what types are allowed (one
is "robust")

Rich

On 7/26/11 3:40 PM, Nyasha Tirivayi wrote:
> Thanks Ariel
> 
> Last time I checked it was not possible to add robust option in the
> logit regression. When I add the bootstrap option, I get confidence
> ntervals and a  z-stat. Yet I thought the t-stat was the important
> indicator of significance. So  which one should I take? If I choose
> the t-stat are you recommending manual calculation?
> 
> Kindly advise
> 
> N.Tirivayi
> Post-doc fellow
> Maastricht University
> Netherlands
> 
> On Tue, Jul 26, 2011 at 6:50 PM, Ariel Linden. DrPH
> <ariel.linden@gmail.com> wrote:
>> As PSMATCH2 uses regression to derive those estimates, you can do the same
>> manually. In fact, I'd suggest you add robust or cluster SE's (cluster on
>> pair ID). PSMATCH2 creates all the variables you'll need to run the
>> regression.
>>
>> Ariel
>>
>> Date: Mon, 25 Jul 2011 23:22:34 +0200
>> From: Nyasha Tirivayi <ntirivayi@gmail.com>
>> Subject: st: Propensity score matching: confidence intervals
>>
>> Dear All
>>
>> I would like to ask for your advice on the psmatch2 software.
>>
>> I am using psmatch2 to determine program impact. I am using the kernel
>> and nearest neighbour estimators. Kindly advise on the following:
>>
>> 1. To report significance of ATT which test statistic should I use,
>> the one labelled T-stat or the z-statistic after bootstrapping?
>> 2. If I use the T-stat, is there a way of obtaining confidence
>> intervals? I need to report the confidence intervals and p-valuesl for
>> a medical journal.
>>
>> Kindly advise
>>
>> N. Tirivayi
>> Post-doc Fellow
>> Maastricht University
>> Netherlands
*
*   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