Date:         Tue, 28 Jul 2009 13:48:03 -0700
Reply-To:     Haris <Karovaldas@GMAIL.COM>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         Haris <Karovaldas@GMAIL.COM>
Organization: http://groups.google.com
Subject:      Re: sample size question when using proc mixed
Comments: To: sas-l@uga.edu
Content-Type: text/plain; charset=ISO-8859-1

Dale is right, you should use all the data you have regardless of
completeness per person and do sensitivity analyses regarding the
pattern of missingness.

You should not be loosing observations from 205 read to 167 used.
What is wrong with the 38 lines that did not get used in your analysis?