What you are used to is the correct thing. Recall the assumptions behind the COMPARE MEANS tests in SPSS and elsewhere. OR Could it be that Var A and B do represent the same characteristic (e.g. "before" and "after" treatment) ? and the cases represent "control" and "treatment" cases?) If so, you need to rearrange the data use the appropriate methods. hope that helps. N.F. What ever

-----Original Message----- From: Elisa Gordon [mailto:elisajillgordon@hotmail.com] Sent: Wednesday, July 23, 2003 6:46 AM To: SPSSX-L@LISTSERV.UGA.EDU Subject: compare means question

Can anyone suggest how I might do the following:

I would like to compare the means of 2 subsets of different cases within 2 different variables. I'm using SPSS v.7 (I know its old). So, just to be clear, I'd like to compare the means of cases 10-20 from variable A and cases 15-30 from variable B.

I'm used to only comparing the means within one variable... so this is entirely new for me.

I'd greatly appreciate any suggestions for going through the steps.

Thank you!

Elisa

Elisa J. Gordon, Ph.D. Assistant Professor of Bioethics and Health Policy Assistant Director of Research Neiswanger Institute for Bioethics and Health Policy Stritch School of Medicine, Bldg. 120, Rm. 280 Loyola University of Chicago 2160 South First Avenue Maywood, IL 60153 Tel: 708-327-9220 Fax: 708-327-9209 Email: egordo1@Lumc.edu Email: elisajillgordon@hotmail.com VISIT THE BIOETHICS INTEREST GROUP(BIG)WEBSITE: http://www.meddean.luc.edu/ssom/depts/bioethics/big/ VISIT THE INSTITUTE FOR BIOETHICS AT LOYOLA WEBSITE: http://bioethics.lumc.edu

_____

Tired of spam? Get advanced junk <http://g.msn.com/8HMRENUS/2734??PS=> mail protection with MSN 8.