2008-03-31

javasaurus: (Default)
2008-03-31 01:22 pm

t-test ANOVA question

If you are comparing two means, use the t-test. If you are comparing multiple means, use ANOVA. fine.

What if you have a control and three test groups, and you want to compare the means of each test group to the control (but not to the other test groups).

A coworker suggests ANOVA followed by Tukey, but I'm thinking multiple t-tests is the way to go here. Opinions?