As far as I can understand this, your problem
sounds like an analysis of variance problem.
It is possible to cycle through t tests for
all distinct pairs of categories, in Stata as in other
software, but it seems
to be generally considered that looking at the problem as
a whole is preferable on a variety of statistical
and scientific grounds.
In any case, doing t-tests separately doesn't make
them independent of each other.
Nick
[email protected]
P chrisman
> Is there a way to do a t-test of the difference in means for an
> unordered categorical variable. I have an unordered
> categorical variable
> and I want to compare the means between group 0 (the reference group)
> and groups 2, 3, 4 individually. I initially thought to do a
> regression
> using the xi: regress command with i.unordered categorical
> variable, but
> I am unsure if this is equivalent to separate t-tests for group 0
> compared to each group 2, 3 and 4 separately. Any
> comments/suggestions would be greatly appreciated.
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
