re: Counsellor constructs
Thu, 10 Aug 95 14:55:00 EST

Whatever one thinks of the Big Five from a constructivist
viewpoint, they are remarkably robust. They arise whether one uses
PCA or different reduction techniques, including content based
approaches, they arise in many languages, whether one examines trait
ratings or questionaire items, and whether one examines self, peer,
spouse, or other types of ratings/judgements. Researchers often
examine more than the 5 factors. The claim is that there is no
replicable, interpretable, independent additional factors, though
trait theorists are not in total agreement (see, eg., Tellegen). PCA
is, of course, rather arbitrary in where it places the axes. In a
certain sense, the Big Five claim in a slighlty weaker sense, is that
the space of inter-individual trait judgements is, at a superordinate
level, 5 dimensional, and that the Big Five factors are a good set of
axes for this space. Of course, one could choose to place the axes
elsewhere, as is done by those into the Interpersonal Circumplex, who
place the Love and Dominance axes at about 45 degrees different, but
marking the same space as, the Big Five Extraversion and Agreeableness
In an examination of therapists construct systems, I showed that
the Big Five presenented a good set of axes for their (rep grid based)
systems, but that the BF were not necessarilly better than a random
collection of their personal constructs. I can't remmember if I
presented this at the Albany or San Antonio PCP conferences. The
question of whether, and to what extent, individuals use the BF in
their construct systems is an important one that needs further
investigation. A colleague and I were going to study this, but he
abandoned the project. If done right, this study would be
extraordinarily important.


Stephen Soldz
Director of Research and Evaluation Department of Psychiatry
Advocates for Human Potential Harvard Medical School
Sudbury, MA
531 Beech St.
Director of Research Roslindale, MA 02131
Boston Institute for Psychotherapy (617) 469-3576
[Sorry folks, I respond in multiple roles!]