I don't have you post in front of me and am going on memory of
it.
As I recall, you want to show that the stick patterns test is not
measuring what it is claimed to measure, namely, something about
spatial ability. You ended your post by asking: Am I being too
ambitious?
What makes you say that it is not measuring what other people say
it is? The different stategies which people use to do the task?
This might be a good application for q methodology.
Write down each of the strategies on an index card. Select a
group of people, perhaps researchers of this task, and ask each
one to do a q-sort as follows: Rank order the strategies from
those which most likely involve spatial ability to those most
unlikely to involve spatial ability. Intercorrelate the q-sorts.
Factor analyze the matrix of correlations. If more than one
factor emerges, you have some support for your position. If only
one factor emerges, you have some lack of support for your
position. If more than one factor emerges, showing that people
have different views on the matter, you could gain some sense of
what each view is by looking at the qsorts for each grouping.
ยทยทยท
To: Pat Alfano
From: David Goldstein
Subject: stick patterns
Date: 08/01/92