Date:         Fri, 17 Jul 1998 13:24:30 -0400
Reply-To:     lloyd@vms.cis.pitt.edu
Sender:       "SPSSX(r) Discussion" <SPSSX-L@UGA.CC.UGA.EDU>
From:         lloyd <lloyd+@PITT.EDU>
Organization: Katz Graduate School
Subject:      calculating freq differences using chi-square..
Content-Type: text/plain; charset=us-ascii

Hello,
 
Can someone verify the following...
 
In SPSS8.0 I used the Statistics-regression-logistic selection to
calculate if the number of subjects completing a task in was
significantly different in three different conditions (thus the
independent variable is treatmentcondition). Thus I have an independent
nominal variable and a dependent dichotomous variable called gotit that
simply contains a 1 if the subject got the task and a 0 if not.
 
After selecting statistics-regresion-logistic, I selected gotit as the
dependent variable and treatmentcondition as the covariate. I then
pressed the categorical button in SPSS and selected treatmentcondition
as a categorical variable. I then pressed OK to run the statistic - to
obtain the chi-square statistic.
 
Is this an appropriate way to investigate significant differences in the
completion rate of a task based on different treatments and/or is there
another way.
 
Thank you for any suggestions...
 
 
 
 
--
 
-------------------------------------
Karl Brett Lloyd
University of Pittsburgh
Katz Graduate School
412-856-8571
lloyd@vms.cis.pitt.edu
-------------------------------------