Re: clustering factor with spss
From: Art Kendall (Art_at_DrKendall.org)
Date: 11/25/04
- Next message: Art Kendall: "statistics & voting in US"
- Previous message: Lady Chatterly: "Re: R/S System: Advanced Programming, San Francisco***December 20-21"
- In reply to: wainer lusoli: "clustering factor with spss"
- Messages sorted by: [ date ] [ thread ]
Date: Thu, 25 Nov 2004 02:44:47 GMT
SPSS has the ability to do many forms of factor analysis including PCA.
it also has CATPCA for factoring categorical variables.
Most clustering algorithms assume the variables are rather independent.
Factor scores from PCA would be orthogonal - very independent so no
problem on that score. In many situations the variables that comprise a
factor are summed into a score with unit weights, 1 if the item loads
highly and cleanly on the factor, and zero otherwise.
SPSS has many similarity measures, and several clustering methods,
including the new twostep procedure which can be used with continuous
and/or categorical variables.
Of course a lot depends on the nature of your variables, and how the
cases were chosen.
Art
Art@DrKendall.org
Social Research Consultants
University Park, MD USA
(301) 864-5570
wainer lusoli wrote:
> Hallo -
>
> Hope this is the ritgh place to ask.
>
> Does anybody know if SPSS supports clustering of cases on basis of PCA
> scores, and how. (before I scratch my nail in vain)
>
> Much appreciated
>
> Kind regards
>
> Wainer
>
> w.lusoli_at_salford.ac.uk
- Next message: Art Kendall: "statistics & voting in US"
- Previous message: Lady Chatterly: "Re: R/S System: Advanced Programming, San Francisco***December 20-21"
- In reply to: wainer lusoli: "clustering factor with spss"
- Messages sorted by: [ date ] [ thread ]
Relevant Pages
|
