Re: Grid analysis
ANDERSTM@ctrvax.Vanderbilt.Edu
Mon, 12 Feb 1996 08:45:26 -0600 (CST)
Dear Tim Connor:
The primary data that you present for INDIVIDUAL clinical interpretation
is confusing to me. This is because 1) I think the Rep Grid makes for
an interesting research tool with large samples, but 2) little can be
gleaned for INDIVIDUAL anlaysis without good qualitative follow-up data
about what the constructs mean to the person, and 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. Using PCA of rep-grids seems to be part of the
"scienticisms" that keeps this originally innovative tool in the stone
ages of science. Without collecting AT LEAST 50 constructs, then why
bother? Instead, why not do what my mentor (Larry Leitner) refers to
as the "What's happening?" test for each of the constructs?
My apologies for the irreverent tone,
Timothy Anderson
NIMH Post-doctoral Fellow
Vanderbilt University
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%