Date:         Wed, 12 Apr 2000 08:14:59 +1000
Reply-To:     Graham Tyson <gtyson@CSU.EDU.AU>
Sender:       "SPSSX(r) Discussion" <SPSSX-L@LISTSERV.UGA.EDU>
From:         Graham Tyson <gtyson@CSU.EDU.AU>
Subject:      unequal cell sizes
Content-Type: multipart/alternative;

I have not had any responses to this message which I posted a couple of
weeks ago, so I thought I would try again.

I want to test for differences between four groups after controlling for a
number of demographic variables. I have used GLM to perform a oneway ANCOVA
to do this. However, the cell sizes are very different with one cell having
about 2,500 observation and the others 100 - 200 (these reflect differences
in the population). Are these differences likely to effect the outcome and
is there a better way to do the analysis? Any references would be good.

Thanks

Graham Tyson

[text/html]