I am posting this to both the personal construct list and to the
Rorschach list as I think it is pertinent to both. I appologise to those
of you who are not familiar with both areas.
There has been some extensive discussion recently on the PCP list
regarding how to measure the constructs of counselors. I found this
highly relevent to a problem which I have been struggling with for many
years and I posted a previous message which brought some interesting
responses. However, several of the respondents wanted to know more
specifically what I was trying to do so that they might be more helpful.
I work extensively with the Rorschach Inkblot Technique. In this
procedure, someone is shown an ambiguous form as asked to say what it
looks like or reminds them of. The key to the technique is that one
person might say it looks like a butterfly and another will say it looks
like a bat or three people dancing. The question I posed early on in my
career was why this was done. Why does one person see one thing and
another sees something else in exactly the same blot?
Those of you are are PCP oriented will probably reply that it depends on
their construction of the world. I quite agree and that is what makes
the Technique so valuable.
Over the years, I have learned that the answer to my deceptively simple
question has to do with how a person structures his or her whole world.
However, I reasoned that perhaps there is some connotative meaning that
is common to all or many of the things seen (which we call percepts) and
I wanted to find out if this was true. In order to do this, it was first
necessary to measure the connotative meaning of the percept to the
subject giving it.
My first attempt used Osgood's semantic differential. It turned out
(quite obviously I learned later) that this was useless since most of the
scales were not relevent to what I was asking about. Later, after
learning about Kelly and the repertory grid, I tried again using a new
set of scales.
I could not use a standard grid technique since I needed to end up with a
set of common scales. To do this, I used a traditional three group
comparison technique. (For those not familiar with PCP, I used typical
Rorschach percepts, exclusively animals and people, and asked a small
sample of subjects to consider three of them and tell me in what way any
two of them were similar but different from the third.) In this way I
developed a series of constructs from the various subjects. I then
informally examined them to determine what dimensions were being used. I
learned the subjects seem to be using five or six bi-polar constructs and
I proceeded to then use a grid to measure the meanings of the percepts to
the subjects giving them. (Again, for the Rorschach people, I asked the
subjects to say if they would describe the bat they saw as "dangerous",
"safe", "powerful", "weak", etc.)
Once again, I ran into severe problems with reliability. On retest, I
got different answers. (Grid was retested, not the Rorschach.)
The ideal solution would be to use a repertory grid and let the subjects
come up with their own constructs. Unfortunately, that doesn't allow me
to compare the meaning (as measured on the subject's own constructs)
given to a particular response with the meaning given to a response given
by another person but scored the same on the Rorschach. In other words,
are there any meaning commonalities among responses that receive similar
scores on the Rorschach. (For the PCP people, Rorschach scores are not
necessarily based on the content of what is seen. They can be based on
such things as whether the percept is seen as moving or whether it is
described as colored.)
Penny Jamison uses a technique very similar to what I used to get the
constructs I used. It is much like the techniques described in looking
for commonalities among the constructs generated by the counselors.
She said:
>>>
When I first go
through a protocol, I look for themes. For example, in one I was doing
yesterday, the major categories seemed at first to be around activity and
passivity (lots of movement in the protocol that seemed unusual to me).
A second obvious theme seemed to be around mutilation--or a sort of
violent incompletness. A third had to do with male vs. female. I
initially began clustering three ways...and noting those response that
were more common (derived from the Exner) as opposed to those that were
highly idiosyncratic. From there I played around with the less common
themes...e.g.natural sett{ngs. It turned out that the gender category
was related to the incompleteness category-- through aggressive, angry
imagery. (And if I get much more involved, I will successfully avoid
getting a report written this morning)...
It is somewhat like an analysis of a night's dream sequence. Most of
the initial work is a sort of free floating attention that allows a
variety of categories to surface. Then I try out the ones that do
surface.
<<<<<
Now to the question (finally)! Anyone have any ideas on how I can
proceed from here? I want to find out if there are any common meanings
underlying responses scored in a similar way on the Rorschach. How do I
measure those meanings.
Thanks for reading this far.
Jack
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%