First off, let me thank you all for your thoughtful replies. Very helpful,
every one of them.
Beverly: Yes, there was some confusion on my part with the usage of the
terms positive and negative in reference to construct poles in the
literature. My query concerns your third usage, "the side of the
construct that is most valued". Thanks for clarifying.
Several of you mentioned the importance of context in the use of the
repertory grid. This was also dealt with in some of the Yorke articles
that were suggested. Note that I am using grids in a (simulated)
vocational counselling setting (the work I am doing here is all
experimental, BTW). With vocational counselling the context is quite often
one of a need to decide among elements (job options). Thus, the grid
represents not simply a construct system, but also an individual's
decision making space.
Marcus says "I've been using PCT and grids in careers guidance since about
1978. I see no reason why one pole of a construct should necessarily be a
'preferred pole' although admittedly clients usually do offer such
negative/positive constructs in practice" This is my experience as well.
When vocational clients are working with a grid, they understand that a
decision looms large, and they seem to want to use the constructs to rate
the value (as opposed to simply the meaning) of the element job options.
Thus they often attach positive and negative values to the construct poles
and expect this to be reflected in rating totals.
This seems to be supported by the literature. For example, in an article
by G. Neimeyer (1989) in the Journal of Counseling and Development, one of
the analyses suggested (among differentiation and integration) is simply
to calculate a sum of ratings for each element. The element rated
positively 67% of the time is seen as a more attractive job option than
one rated positively only 33% of the time. There are other examples of
this.
I am aware that the grid was originally proposed by Kelly for the
construing of various roles in a person's life, and rating for the purpose
of determining the most favored person was never the intention. But as
the grid is adapted for other purposes (i.e., decision making or
representing a decision making space) the method must adapt with it. If we
are going to sum ratings per element as above then we must be sure
that a positive rating for each cell is indeed something valued by the
client. This is why I was encouraging my participants to pick one pole as
positive and the other negative, relative to each other.
Which brings me back to my original problem: What to do when constructs
are elicited that are opposite in meaning, but similar in value? For
example, works indoors -v- works outdoors, both poles of which were seen
as positive for one of my participants. What can be done to ensure that
constructs have a positive pole and a negative pole so that positive
ratings are truly valued, and negative ratings are not valued by a grid
user?
John has called my attention to the section "Splitting Constructs" from
the Enquire Within site. I actually have done this before, with some
success. Nice to see some corroboration.
Another way to solve the problem is simply to side-step opposing poles and
substitute in concepts for constructs. For example, get participants to
nominate important "aspects" (not constructs) of occupations by free
elicitation, and then rate probability of the aspect. Thus, the
aspect "job security" would be rated positively with higher guarantees of
this aspect, and lower (into the negative numbers) with lower guarantees
of this aspect. PCT purists will no doubt be horrified at this
suggestion, and I myself don't see it as the best means for representing
an individual's construing. However, I did try it with a few willing
participants and it was very easy to administer. As well, the
participants found the rating this way intuitively straightforward, and
responded positively to the results.
Another way to ensure that constructs are elicited in terms of value is
simply to change the wording of the method of triads instructions to
reflect the importance of constructs in terms of their value.
Essentially,
"In what way are two of these the same in value, yet opposite in value to
a third"
This reminds the participant with each construct elicitation that the
context for rating the elements is in terms of values, and not simply
meaning. I was hoping that someone would have seen something like this
before. Failing that, I am wondering what all of you think of this
change.
In many ways, I am learning as I go along, so please keep your feedback
coming.
-- R
P.S. Bob, your reference to Mackay, can you give me a date and a journal?
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%