Repeated measures MANOVA (Within-Subjects Variables)

Cameron Norman

unread,

Sep 3, 2003, 4:00:27 PM9/3/03

to

I am a novice to the repeated measures MANOVA and am trying to analyse data
I have collected at baseline, post-test and 3-month follow-up (i.e., 3
levels). I am using SPSS 10.1 for analyses and am having problems once I get
to defining my within-subjects variables after defining my factors. I have
more 'slots' than variables.

For the within-subjects variables I have two factors with three levels each.
When I click on 'Define' I see 9 'slots' for my within-subjects variables.
Yet, I have only six variables (factor 1 measured at baseline, post-test &
3-month AND factor 2 measured at the same time).

My apologies to the list if this question seems very elementary, but I am
genuninely baffled by this and am suspecting that I need to calculate some
new variables based on each combination (1,1 1,2 1,3 2,1 2,2 2,3 3,1
3,2 3,3) in the nine 'slots'.

Many thanks to anyone who can shed light on this.

Cameron

Any help on this would be greatly appreciated as my local resources are
simply not familiar with MANOVA's performed at the multi

Kylie

unread,

Sep 3, 2003, 11:34:58 PM9/3/03

to

Hi Cameron,

Sounds like you've got what SPSS calls a doubly multivariate repeated
measures design.

To set it up, in the first 'Define Factor' dialogue box of GLM Repated
Measures define your within-in subjects 'time' variable with 3 levels
and then click the Measure button. Enter a name for the 2 dependent
variables, clicking Add after each one, and then Define. The main
Repeated Measures dialogue box will now have six variable slots -
three time points for each of your two dependents. Any
between-subjects factors or covariates can also be added as per
normal.

If you have access to the manuals there is an example in SPSS Advanced
Models 10.0, chapter 13, GLM Repeated Measures Examples, example 2
using the 'new drug.sav' data file.

HTH,
Kylie Lange.

"Cameron Norman" <cameron...@utoronto.ca> wrote in message news:<HKnM6...@campus-news-reading.utoronto.ca>...

Timothy W. Victor

unread,

Sep 4, 2003, 12:27:52 AM9/4/03

to

I would recommend having a look at Tabachnick and Fidell's, _Using
Multivariate Statistics_
(http://www.abacon.com/tabachnick/stats/index.html). They do an
excellent job of describing how to use the software for this design.

jim clark

unread,

Sep 8, 2003, 10:08:57 PM9/8/03

to

Hi

On Wed, 3 Sep 2003, Cameron Norman wrote:

> I am a novice to the repeated measures MANOVA and am trying to analyse data
> I have collected at baseline, post-test and 3-month follow-up (i.e., 3
> levels). I am using SPSS 10.1 for analyses and am having problems once I get
> to defining my within-subjects variables after defining my factors. I have
> more 'slots' than variables.
>
> For the within-subjects variables I have two factors with three levels each.
> When I click on 'Define' I see 9 'slots' for my within-subjects variables.
> Yet, I have only six variables (factor 1 measured at baseline, post-test &
> 3-month AND factor 2 measured at the same time).

I'll use measure to refer to what you have called factor 1 and
factor 2, so that we can use the term factor to describe the
design. Your design has two factors, one with two levels
(measure 1 and measure 2) and one with three levels (baseline,
post, follow). 2x3 = 6 slots to be filled.

Best wishes
Jim

============================================================================
James M. Clark (204) 786-9757
Department of Psychology (204) 774-4134 Fax
University of Winnipeg 4L05D
Winnipeg, Manitoba R3B 2E9 cl...@uwinnipeg.ca
CANADA http://www.uwinnipeg.ca/~clark
============================================================================