Date:         Sun, 22 Mar 2009 08:20:29 -0500
Reply-To:     Phil Schubert <schubert@acu.edu>
Sender:       "SPSSX(r) Discussion" <SPSSX-L@LISTSERV.UGA.EDU>
From:         Phil Schubert <schubert@acu.edu>
Subject:      split plot or mixed between-within ANOVA
Content-Type: multipart/alternative;

In a case where a mixed between-within ANOVA seems like the right
statistical tool, does anyone know if its possible to use this tool with
several "within" variables in the same model?
For example, suppose i have measures for 10 different continuous variables
at time 1, 2 and 3.  These variables represent my "within-subjects"
variables and time represents my "within-subjects factor."  Next, I have a
categorical grouping variable representing the "between subjects" variable.
 Am i confined to testing one continuous variable (with 3 measurements) at a
time or is there some way to assess the three measurement points among all
10 variables at once?

I guess i am looking for the "within-between" equivalent of a multivariate
anova that allows assessing the significance of several dependent variables
at once; only in this case, i have multiple time measurements for each
dependent variable.  Otherwise, I am concerned that executing 10
within-between anovas dilutes my alpha value.

Any ideas or insight would be greatly appreciated!

--
Phil Schubert
Abilene Christian University
Executive Vice President
ACU Box 29120
Abilene, Texas 79699-9120
325/674-2795

[text/html]