continuous variables in MANOVA
Don Burrill
> I have run an MANOVA and an MANCOVA on data involving:
> 3 continuous dependent variables
> 2 continuous independent variables
> 1 discrete variable (two levels)
> One of my committee members insists that I cannot do a MANOVA with
> 2 continuous independent variables. ... I need to ask two things:
> 1) Can I do MANOVAs with these variables ?
> 2) If you think I can, could you cite me references of studies or
> books that have done this or recommend it. And if you think I
> cannot, tell me why specifically.
---------------------------------------
You did not specify the model you are trying to fit.
At face value, and assuming the one discrete variable is an independent
variable, short answers are:
1) No. If univariate ANOVAs were not appropriate for addressing
your dependent variables singly, MANOVA is not appropriate.
2) In MANOVA (as in univariate ANOVA) the independent variable(s)
are categorical (or "discrete"), not (quasi-)continuous.
A more detailed answer will be sent privately, not to clutter up the public
list-serve space. (If anyone else wants a copy, please ask.)
-----------------------------------------------------------------------------
Donald F. Burrill (416) 923-6641 ext 2460
The Ontario Institute for Studies in Education DBur...@UTorOISE.bitnet
Toronto, Canada M5S 1V6 DBur...@OISE.UToronto.CA
-----------------------------------------------------------------------------
David Nichols
As with the above comment, more specification needs to be given as to
exactly the model and research questions involved, but the statement that
you can't use continuous variables or more than one is incorrect. Some
people define ANOVA so that there are only categorical independent
variables (e.g., Scheffe), but the underlying model for ANOVA/MANOVA,
ANCOVA/MANCOVA and regressions is the same, and the status of the predictor
variables comes into play only for interpretation and parameterization of
models. All are well defined. Further followup questions/comments welcome
at the address below or on the wire.
--
David Nichols Senior Statistical Support Specialist SPSS, Inc.
Phone: (312) 329-3684 Internet: nic...@spss.com Fax: (312) 329-3657
*******************************************************************************
Any correlation between my views and those of SPSS is strictly due to chance.