Date: Mon, 12 Dec 2005 10:01:39 -0600
Reply-To: "Swank, Paul R" <Paul.R.Swank@uth.tmc.edu>
Sender: "SPSSX(r) Discussion" <SPSSX-L@LISTSERV.UGA.EDU>
From: "Swank, Paul R" <Paul.R.Swank@uth.tmc.edu>
Subject: Re: mixed design repeated measures question
Content-Type: text/plain; charset="us-ascii"
What it means is that the difference between the pre- and posttests for
the control and treatment groups is not the same for the two levels of
SES. You could do simple main effects because it is possible for the
differences to be significant in both cases but to be larger at one
level that the other. However, I usually just graph the results. With 1
df for the test, all independnet hypotheses are already tested. Simple
man ewffects do represent confounded effects.
Paul R. Swank, Ph.D.
Professor, Developmental Pediatrics
Director of Research, Center for Improving the Readiness of Children for
Learning and Education (C.I.R.C.L.E.)
UT Health Science Center at Houston
From: SPSSX(r) Discussion [mailto:SPSSX-L@LISTSERV.UGA.EDU] On Behalf Of
Sent: Monday, December 12, 2005 12:58 AM
Subject: mixed design repeated measures question
I'd like to ask your help in interpreting findings on a mixed-design
repeated measures test.
In my design I have one within factor (pre-post test scores- I named
this factor as Time) and 2 between factors (ses and treatment group). My
analysis gives a significant interaction of all these three factors such
Interaction of Time, SES, and Treatment is F (1,229) = 6,505 p=,011.
As far as I understand this means: The intervention's effect for each
treatment group (exp vs control) is dependent on the SES (hi-low).
How do I conduct further analysis to understand where the difference
exactly comes from? Graphs show that my treatment has a differential
effect between my low- hgh ses groups. What is the part of the analysis
I need to look at for this effect? (contrast subcommand only gives me
the main effects for treatment and ses)
Is it accceptable to split the whole sample by SES and Treatment and do
a paired t-test where I compare the average post-pre scores of
experimental / control groups in each SES group?