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]
st: how to test DD assumption about the similar trend before treatment
From
"Li, Rui (CDC/ONDIEH/NCCDPHP)" <[email protected]>
To
"[email protected]" <[email protected]>
Subject
st: how to test DD assumption about the similar trend before treatment
Date
Wed, 19 Oct 2011 20:13:02 +0000
Hi,
I am brand new to the group. I did check the FAQs. However If I ask questions being answered before, please forgive me.
I am using a DD model and testing whether the treatment group and control group have similar trend before the policy changed (repeated cross-sectional data). I have at least 5 years of pre-policy data. I tested by the following two ways:
1. Add a year trend and interaction between year and treatment group, then if the interaction terms was not significant, I concluded that the two groups had similar trend before the policy change (like a linear spline analysis with the year before policy change as the knot).
2. Run separate regressions for treatment group and control group using only pre-policy data, I coded the year variables as categorically data because the time trend might not be linear. Then I compared each coefficient for the year variables in the two regression analysis using ttest.
Are these the right ways to test the assumption?
Thanks!
Rui
*
* 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/