Re: st: Difference-in-Difference with N=4, missing std. errors

[Steven Samuels <sjsamuels@gmail.com>]

A make-shift solution is to enter the original eight observations into  
-reg- with an added country indicator. This will give  standard errors  
with three degrees of freedom.



On Dec 5, 2010, at 11:37 AM, Stas Kolenikov wrote:

That's a theory of linear models issue: you have 4 equations with 4
unknowns, so you consume all information in the sample; you don't have
any degrees of freedom to estimate variance of the error term in the
regression, and hence end up with no standard errors or F-tests.

On Sun, Dec 5, 2010 at 10:10 AM, Jennifer Maki
<jenanneharris@hotmail.com> wrote:
> Statalist:I'm trying to use difference in differences to tease out  
> the effect of a policy change.  I'm using country level data and  
> have yearly figures for both countries.  I averaged them so that I  
> have a figure representing the 5-year average pre & post treatment  
> for both countries. leaving me with 4 observations.  When I estimate  
> this using the regress command in Stata, the coefficient on the  
> interaction term is what I get when I calculate the difference in  
> difference by hand, but I don't get standard errors or p-values in  
> the output.  Is there a way to get this information so that I  can  
> present the statistical significance of my findings?
> Thanks you,Jen
> *
> *   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/



--
Stas Kolenikov, also found at http://stas.kolenikov.name
Small print: I use this email account for mailing lists only.

*
*   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/