Re using grids to assess change produced by a non-PCP intervention

Tony Downing (
Tue, 1 Jun 1999 00:00:33 +0100

In response to the friendly, helpful radical challenges to the project
we're embarking on, that Devi and Barry Cooper have recently posted (in
addition to having given much direct help - thanks Devi!), I'd like to
speak in defence of combining the study of personal constructs with the
dreaded "positivistic" methods. As you'll see, I'm not, as yet, anyway, a
true believer in PCP but I think it has a lot going for it. This message
has turned out longish, but I think the issue is a good example for
considering how purely Kellyan the PCP approach can be/has to be/needs to
be - so read it if you are interested!

The project we are embarking on is: assessing change produced by a training
course for parents whose young children have severe communication
difficulties. (At least half of them will be autistic.) It teaches the
parents to recognise incipient communication in their children and to
encourage it. The speech & language therapists, in this particular
programme, work with the parents - not directly with the children - but
they video the parents with their children and the videos are discussed
individually and with the other parents in the group.

Informal reports indicate great success, in improving the communication of
the children (the direct target aim) and in reducing the stress of the
parents, and, it seems, in helping these parents, often extremely upset for
very good reason, to be happier with and about their children. There is
some research, to which I dare say many members of this forum would apply
the swearword "positivistic", which has shown quantitatively some good
evidence to say that this method does work. This has not been done because
the researchers wanted to ride roughshod over individual uniqueness, but it
just was important to answer the question "Does it work?". This is a
matter of great practical importance, as it feeds into such decisions as
whether to divert resources from attempting conventional speech and
language therapy with these children, towards running training courses such
as the one I've outlined, about which the parents are so enthusiastic.

The little study we are proposing to do aims to find out, in individual
detail, the nature of the cognitive changes that take place in the parents
as the course proceeds - and IF similar stories come from many individuals,
then also to draw out some generalisations. Pretty clearly, what goes on
involves change in how the parents construe their children - at a broad
level, and also in fine detail, such as how the parents construe the
meanings of individual pieces of behaviour on the part of their children.
This does look like a job for Rep. Grids and PCP, doesn't it?

Thus, we are hoping to assess changes in how the parents construe their
children's behaviour, at the end, compared to the beginning of the course.
We will be interested in contents, and also in abstract measures such as
cognitive complexity, which we think the course should increase, and
tightness of construing, which it surely should help to reduce.

In this context, I don't think we can follow Barry Cooper's suggestion:

>But why not accept the rep grid as an intervention or part of the training?

The therapists running the parent interaction course do what they do, and
we have the opportunity to try to find out more about how it works. What
they do seems to work, and if in the light of more understanding of how it
does so , and how well, they can maintain or increase the support of their
Health Service Trust's Speech & Language Department, they will continue to
do it, perhaps even do more of it. And, just as important,they may modify
what they do in the light of this research, depending on what it shows. So
I hope I make the case that finding out how this therapy works is important
and worthwhile.

It is true, that there is a possibility that doing PCP therapy instead of
doing what they do would also produce good effects, but at the moment they
are not in the market for that, they want to know how and how well what
they ARE doing works.. And after the forthcoming course is over - and our
research project is over too, their courses will be run without the rep
grid element, so it's not relevant right now, from the point of view of
service provision, to know how good it would be to have a combination of
what they presently do with PCP therapy. Thus, there is a legitimate need
to investigate the cognitive changes that the course produces, while trying
to minimise the effects of our observation method. Of course we can't
minimise it to zero, but surely it's best not to maximise it! I am, I
suppose, wanting the grids to be "a test", but not one where we closely
prescribe all the answers. And we will have a control group, who, this
time round at least, will receive standard speech and language therapy,
rather than the parent interaction course, so that should help to assess
the course indepently of the rep. grid work that we will bring to it.

My earnest messages over the past week or so, asking how bad it will be if
we don't check out the interpretations of the parents' grids with them at
the time, spring from that dilemma. Of course if we are likely to
seriously misinterpret their construing, then perhaps we have to check out
the interpretations of their grids with them - but it will carry a risk.
For that reason, I'd sooner leave checking things out until after the
second grid, at the end of the course. That way it would not be likely to
have such a strong influence on the before/after comparison. I certainly
don't want to lose the richness of the parents'' comments and answers,
discussing their grids, so in the circumstances, maybe that's the best
answer? I know there would be a risk that they would not remember very
well how they were thinking 3 months back, when doing their first grid, but
I think I've just talked myself into deciding that that compromise plan is
what we'll do.

Comments? Hell-fire? Exorcisms?

Tony Downing,
Dept. of Psychology, University of Newcastle upon Tyne, England.