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: DiD
From
[email protected]
To
[email protected]
Subject
Re: st: DiD
Date
Thu, 31 Mar 2011 16:32:33 +0200
> Hello Federico,
> You are correct, treat and treat_post05 are highly
> correlated.Should I drop "treat"?
I don't think it would make sense. This variable should control for differences between treated and non-treated that are already there before the implementation of the program. Thus, excluding it would increase bias.
When you say your dataset is unbalanced, what do you mean? You mean (non-) treated are too many (few), or that you have some individuals that are observed only in 2000 (2007)? Because if correlation between treat and treat_post05 is very high, may be you have too many treated people that are observed only in year 2007.
>
> Indeed, I also created an interaction dummy between
> treat_post05 and a covariate (X1). Hence, the full model is:
>
> regress y treat post05 treat_post05 treat_post05*X1 X2 X3
Can you tell me what kind of variable X1 is? Is it binary? Are X2 and X3 highly correlated with treat_post05?
Federico
*
* 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/