Possibly the dumbest question ever

Home > Statistics Every Writer Should Know > The Stats Board > Discusssion

Possibly the dumbest question ever
Message posted by Holly on November 28, 2000 at 12:00 AM (ET)

A very simple question: I am trying to determine whether three different groups differ in terms of whether or not they have been diagnosed with depression (coded as "1") or not (coded as "0"). What test will answer this question for me? That is, what statistic allows me to determine whether two or more groups differ on a dichotomous variable?
Thanks!

READERS RESPOND:
(In chronological order. Most recent at the bottom.)

Re: Possibly the dumbest question ever
Message posted by Phil on November 29, 2000 at 12:00 AM (ET)

I suggest you start with the Chi-Square Test for independence. If you have a 2x3 matrix of group vs. diagnosis with counts in each box in the matrix, that should show you if there is a significant difference among groups.

Your $5 contribution helps cover part the $500 annual cost of keeping this site online.