Checking out the grid with the subject

Tony Downing (
Sun, 30 May 1999 12:59:57 +0100

Devi raises the profound point about rep grids, that:

>a high matching score between any two elements, or constructs, is just a
> number representing a rating. If you want to draw inferences about about
> a person's structure in general, rather than simply as shown by the
>ratings, it's often faster to ask the person. Particularly, a high
>matching score between two constructs may _not_ mean that the constructs
>are causally linked ("whenever you think a - not a, you also tend to
>think b - not b; does that mean that b - not b is implicationally
>dependent on a - not a?"), but simply associated without cause; or,
>indeed, simply coincidentally present, a function of how adequately you'd
>sampled the whole realm of discourse in choosing the elements you used.

This problem, that the specific instances that you sample, with some
purpose in mind, always comes bundled with a lot of particular features
which are not what you're after, is a special case of the general research
design problem of confounding with extraneous variables. In most kinds of
research, it's dealt with by having samples big enough so that all these
particular but irrelevant characteristics ("error variance") tend to cancel
out. In rep grid work it couold in principle be dealt with, presumbly, by
having gigantic grid, so that each _kind_ of element that you'd be
interested in would be represented by a lot of _particular_ elements - but
presumably that is just not practicable.

In that case, Devi's recommendation, that we check out with the
subject/client that small element distances or high construct correlations
are not just arising by chance because of unforseen quirks of the
particular choice of element, seem very important.

When it comes to plans, such as ours, to use rep. grids in investigate, in
as idiographic a way as possible, the nature of the cognitive changes that
we hope will be produced by a training course, there is a possible
problem. The rep grid procedure itself may well do some good! This is
fine for the individual but not fine for assessing the ways in which the
training, rather than the rep. grid, affects their construct system. In
this particular study, we don't want to find ourselves having evaluated the
effect of repertory grid therapy! Presmably, if we discuss the way the
analysis comes out with the participants immediately (rather than in
sessions after the end of the course) the risk of the rep. grid experience
itself producing change is much greater.

Does this mean it's not such a bright idea after all, to use rep. grid
methods to investigate these changes?

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