Date:         Thu, 9 Jun 2005 16:03:05 +0100
Reply-To:     Jeremy Miles <jnvm1@york.ac.uk>
Sender:       "SPSSX(r) Discussion" <SPSSX-L@LISTSERV.UGA.EDU>
From:         Jeremy Miles <jnvm1@york.ac.uk>
Subject:      Power from MANOVA
Content-Type: text/plain; charset=ISO-8859-1; format=flowed

Hi All,

I am trying to use the MANOVA approach to estimate power, using SPSS 12,
which is outlied in the paper D'Amico, E. J., T. B. Neilands and R.
Zambarano (2001). Power analysis for multivariate and repeated measures
designs: a flexible approach using the SPSS MANOVA procedure. Behavior
Research, methods, instruments and computers 33(4): 479-484.

The analysis is pretty simple.  Baseline and outcome measurements, two
groups, test for difference using baseline measure as covariate.



However, it ain't working.  It works fine when I don't have a covariate
in, but when I put in the covariate, it all goes wrong.  Specifically it
says that the effect of the group is zero (p = 1) and the power = 0.05.

Here's my code.  Does anyone have any idea what's gone wrong?

MATRIX DATA variables = group rowtype_ baseline outcome
/FACTOR = group
/FORMAT = lower nodiagonal.
BEGIN DATA.
1 mean 0 0
1 n 400 400
2 mean 0 0.3
2 n 400 400
. sd 1 1
. corr 0.45
END DATA.


MANOVA outcome by group(1,2) with baseline
/METHOD UNIQUE
/ERROR WITHIN+RESIDUAL
/MATRIX = in (*)
/POWER T (0.05) F(0.05)
/PRINT SIGNIF (MULT AVERF)
/PRINT PARAM (ESTIM) .

Thanks,

Jeremy




--


Jeremy Miles
mailto:jnvm1@york.ac.uk             http://www-users.york.ac.uk/~jnvm1/
Dept of Health Sciences (Area 4),  University of York,  York,  YO10 5DD
Phone: 01904 321375       Mobile: 07941 228018         Fax 01904 321320