Re: Grid analysis

Rainer Riemann (upsyf077@HRZ.Uni-Bielefeld.DE)
Fri, 16 Feb 1996 14:35:20 MEZ

Dear Dr. Anderson,

in your recent posting you state:
> 3) the use of principal
> components, a common practice with gridders, is absurd (in my not so
> humble opinion) since the so-called sample size is necessarily too small
> for factor analysis.

Although stated frequently, the above statement is simply false. At
best it reflects a basic misunderstanding of the use of principal
components analysis for the analysis of grids.

Recently W. Chambers posted a 'mandala grid'. If you are not among the
happy few people, who can see the basic structure underlying this grid
without an technical tools, do a PCA and you may be surprised what PCA
can do for your understanding of relations among elements and constructs
and between elements and constructs.

Kind regards
Rainer Riemann


PD Dr. Rainer Riemann Phone: <0>521 106 4529 Universitaet Bielefeld Fax: <0>521 106 5844 Fakultaet fuer Psychologie e-mail: und Sportwissenschaft Postfach 100131 D-33501 Bielefeld _______________________________________________________________________________