RE: st: Propensity score matching in stata

["KASHEFIPOUR E." <E.KashefiPour@swansea.ac.uk>]

Thanks a lot Shikha. Your comments were really helpful.
Does the command that you gave me "logit debt size industry_type product_type" match samples by firms and industry? What does this command do?

Thanks


-----Original Message-----
From: owner-statalist@hsphsun2.harvard.edu on behalf of Shikha Sinha
Sent: Tue 24/07/2012 01:52
To: statalist@hsphsun2.harvard.edu
Subject: Re: st: Propensity score matching in stata
 
If you want to estimate the average treatment effects (ATE or ATT),
you can use psm. In your case, are you also looking at the impact of
issuing debt on some outcome(?)? if not then, matching is not a good
strategy and simple probit or logit will solve your problem.

logit debt size industry_type product_type

This model will answer question about which type of company is more
likely to issue debt. Hope it helps.

Shikha

On Sat, Jul 21, 2012 at 2:29 PM, KASHEFIPOUR E.
<E.KashefiPour@swansea.ac.uk> wrote:
> Dear Marco,
>
> Many thanks for your help.
>
> Actually, I am a new user of PSM, and therefore I appreciate if you could let me know about att in more details.
>
> Best,
> Eilnaz
>
>
> -----Original Message-----
> From: owner-statalist@hsphsun2.harvard.edu on behalf of Marco Ventura
> Sent: Sat 21/07/2012 09:21
> To: statalist@hsphsun2.harvard.edu
> Subject: R: st: Propensity score matching in stata
>
> Dear friend,
> The propensity score matching estimator is aimed at working out the average treatment effect for the treated, att, and in its first step it works out the probability of being treated, ie of having issued debt in your case. Therefore, unless you are interested in the att you do not need to use PSM, you may simply run a logit or probit reg. Hope this helps.
> Regards marco
> ----- Messaggio originale -----
> Da: KASHEFIPOUR E. <E.KashefiPour@swansea.ac.uk>
> A: statalist@hsphsun2.harvard.edu
> Inviato: Fri, 20 Jul 2012 18:23:22 +0200 (CEST)
> Oggetto: st: Propensity score matching in stata
>
>
>
> Hi all,
>
> I have a concern regarding the propensity score matching.
>
> I have a list of companies that issue debt between 2007 and 2011. My aim
> is to match those companies with those companies which have not issued
> debt between 2007 and 2011 and investigate the probability of issuing
> debt.
> Can anyone tell me if the propensity score matching is an appropriate
> estimation in my case? If so, which command I can use in stata to match
> those companies with industry and size?
>
> Best,
> ELN
>
> *
> * 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/
>
> *
> *   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/
>

*
*   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/

<<winmail.dat>>