s-news
[Top] [All Lists]

MISSING OBS IN REPEATED MEASURE DESIGN

To: s-news@lists.biostat.wustl.edu
Subject: MISSING OBS IN REPEATED MEASURE DESIGN
From: duo wan <duo_wan@yahoo.com>
Date: Wed, 12 May 2004 20:56:51 -0700 (PDT)

Hi,

 

I have a data set consisting of 3 treatment groups and 1 control group (10 subjects per group); at each of 8 time points, tumor sizes were measured. However, subjects in control group were dying at last 3 time points and none survived in the last time point. Since there are no observations for control group in the last time point, is it reasonable to drop all the observations in the last time point or do some kind of imputation?  The main interest is to find out if 3 treatments can reduce tumor size.

 

Any suggestion will be appreciated.

 

Vincent wan


Do you Yahoo!?
Yahoo! Movies - Buy advance tickets for 'Shrek 2'
<Prev in Thread] Current Thread [Next in Thread>
  • MISSING OBS IN REPEATED MEASURE DESIGN, duo wan <=