Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.


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

Re: st: RV: Doubt about level option in pscore.ado


From   Jorge Eduardo Pérez Pérez <jorge_perez@brown.edu>
To   "statalist@hsphsun2.harvard.edu" <statalist@hsphsun2.harvard.edu>
Subject   Re: st: RV: Doubt about level option in pscore.ado
Date   Thu, 5 Dec 2013 14:23:19 -0500

Standard significance levels are 0.1, 0.05 and 0.01 and you shouldn't
mess with them. Decreasing them is not improving the balance, it's
making it harder to reject the null hypothesis of no difference in
means. So if you get balance after decreasing the level, nothing in
your sample has changed, you just made it harder to reject the null of
balance.

In the previous threads Melisa Garrido and me pointed out that t-tests
are not the best measure of balance. Again check the paper by Kosuke
Imai I linked to in that thread.
--------------------------------------------
Jorge Eduardo Pérez Pérez
Graduate Student
Department of Economics
Brown University


On Thu, Dec 5, 2013 at 10:09 AM, Carlos Tendilla González
<carlos.tendilla@imss.gob.mx> wrote:
> Hi,
>
> Can anyone help me please?
>
> Regards,
>
> -----Mensaje original-----
> De: Carlos Tendilla González
> Enviado el: miércoles, 04 de diciembre de 2013 10:23 a.m.
> Para: 'statalist@hsphsun2.harvard.edu'
> Asunto: Doubt about level option in pscore.ado
>
> Dear Statalist,
>
> I am working with pscore.ado, a user written command from Becker and Ichino.
>
> I want to know exactly what happens when you change the level option, like this:
>
> . pscore totalformal familiar casado edad edad2 hombre escolaridad sector1 sector2 sector3 sector4 sector5 sector6 sector7 sector8 sector9 sector10 sector11 sector12, pscore(ps) blockid(block) level(.1) numblo(10)
>
> . pscore totalformal familiar casado edad edad2 hombre escolaridad sector1 sector2 sector3 sector4 sector5 sector6 sector7 sector8 sector9 sector10 sector11 sector12, pscore(ps) blockid(block) level(.00000001) numblo(10)
>
> I found that when you decrease the level, the balance is better. I think this option decreases the significance level for the p-value of the t-test for equality of means in control and treatment group. So, I should not decrease to much this level. If what I think is right, what is the optimal level of significance?
>
> Regards,
>
> Email secured by Check Point
>
> *
> *   For searches and help try:
> *   http://www.stata.com/help.cgi?search
> *   http://www.stata.com/support/faqs/resources/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/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/


© Copyright 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index