Re: pooled within groups correlation matrix.

[David Marso]

Talino wrote:

>>I was reading an article that suggested that "pooled within-group
correlations among DVs" should be reported when writing a results section
for a MANOVA analysis. The author states that these correlations can be

found in the "pooled within-group correlations matrix" I can't seem to find
the option to report correlation in the MANOVA analysis. Where and/or how
can I generate such a matrix? Is it simple correlation matrix of my two

DVs?

Is it simple correlation matrix of my two
DVs?  

No, it is not!

Any of the following will do !.

DISCRIMINANT
  /GROUPS=region(1 3)
  /VARIABLES=educ paeduc maeduc speduc
  /ANALYSIS ALL
  /PRIORS EQUAL
  /STATISTICS=CORR
  /CLASSIFY=NONMISSING POOLED.

MANOVA educ paeduc maeduc speduc BY region (1,3)  / PRINT =ERROR (COR) .


GLM educ paeduc maeduc speduc BY region
  /METHOD=SSTYPE(3)
  /INTERCEPT=INCLUDE
  /PRINT=DESCRIPTIVE RSSCP
  /CRITERIA=ALPHA(.05)
  /DESIGN= region.

Regards, David Marso