Thinking of what your average non-PCP committee member may want, some
combination of provided and personal items seems appropriate. As I
suggested earlier, defining a domain and generating items is one way
of doing the 'provided' part. Another way is to have a set of
situations that they can do paired comparisons on. Obtain a
similarity matrix and do some ALSCAL to get at their individual
spaces. There may be more or less contradiction among different
people, giving them different 'weirdness coefficients' (weirdness in a
purely statistical sense, of course). *This* may well be related to
the degree of order in their construct systems as shown by a grid that
uses personally-generated situations from their own experience. In
fact, I'd be surprised if there *wasn't* a relationship. Showing a
mapping between personally- and researcher-provided patterns across an
older "conservative" technique and a grid would be a coup for PCP!
> Both asking for a definition and giving some choice seems
> important, so not to impose my construct of risk.
What's wrong with the question "to what extent does this group agree
with my construct system for risky things?" ? Your examples
<snipped> seem pretty good, although a couple of focus groups wouldn't
hurt in refining them.
cheers,
((((((((((((((((((((((((((((((((((((((((((((((((((((((((((
Travis Gee () tgee@superior.carleton.ca ()
() () ()()()()
() () ()
() ()()()()()()()()()
"In science, the more we know the more extensive the
contact with nescience." -Spencer
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%