Reply to Tim Connor

15 Mar 1995 08:38:28 -0500

>From Tim Connor:
> A question re principal component analysis of grids (I'm no
doubt > revealing my statistical naivete, but what the hell): are
the factors > conceptualized as unverbalized superordinate

While knowing the context from which your comments were extracted
would be useful, I will hazard a comment. I still have trouble
understanding how probability works, do you add or multiply all
those coloured balls to end up with 3 red and 1 white? Anyway,
why not look up some of the SAGE series on analysis procedures
or check the book by Hair J, et al. Multivariate Data Analysis,
which is a useful introduction.

My impression is that 'unverbalized superordinate constructs' are
an interpretation (name) given to an extracted component...the
constructs which loaded (correlated highly) with the factor may
have been considered to have all been unverbalised constructs.
Procedures such as pca, mds, cluster analysis are utilised to
reveal underlying structure in the data. This is done by
statistical means. The labels given to this structure is an
interpretation/construction based on a researcher's hunches,
participant feedback or validated by statistical means. The
derived structure is a representation rather than something which
is. I hope this is of some help,
Bob Green.