Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
| From | federico.tedeschi@univr.it |
| To | statalist@hsphsun2.harvard.edu |
| Subject | Re: st: test for independence for variables adding to 100% |
| Date | Fri, 15 Apr 2011 10:13:46 +0200 |
I think a test for independence doesn't make sense in your case: those variables are dependent by construction. However, you could evaluate how a variable affects the relative weight of the other two. To explain what I mean, being q1, q2 and q3 the rate of Elementary, Middle and High Schoolers respectively, you could see how q3 affects q1/(q1+q2). Federico ----- Messaggio originale ----- Da: "Lim, Raymond" <rl2240@columbia.edu> Data: Thursday, April 14, 2011 18:11 Oggetto: st: test for independence for variables adding to 100% A: "statalist@hsphsun2.harvard.edu" <statalist@hsphsun2.harvard.edu> > Hello Statalisters, > Say you have teams composed of 3 types of people: elementary > schoolers, middle schoolers, high schoolers. How do you test > whether teams with a high percentage of high schooler also have > a high percentage of elementary schoolers? The tricky part is > the three must sum to 100% so having a high percent of one group > automatically means having a low percent of the other groups. > > Thanks, > -Raymond > > * > * 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/ * * 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/