Clara,
This is only a partial answer because I am not an expert on how MI works
in Stata, but the equivalent to -ttest var, by(dummy)- without MI would be
-regress var dummy- (your standard bivariate OLS regression). The
coefficient on dummy is the mean difference and the t-ratio and two-tailed
p-value are identical to those produced by -ttest-.
Perhaps another Statalister can confirm or deny that this equivalence
holds in the MI world as well.
David
--
David Radwin // [email protected]
Research Associate
MPR Associates, Inc.
(510) 849-4942
http://mprinc.com
-----Original Message-----
From: [email protected]
[mailto:[email protected]] On Behalf Of Clara Barata
Sent: Wednesday, November 04, 2009 10:42 AM
To: [email protected]
Subject: st: t-test comparing the means of two samples in imputed datasets
Hey!
Any idea on how to apply a ttest to compare means in datasets imputed
with MI (stata 11)?
What would be the equivalent to: "ttest var , by (dummy)" in the MI
world?
I searched through "mi estimate" and "mi test", but couldn't find
anything that matched.
Thanks, clara
*
* 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/
