The Hamburg short psychotherapy comparison experiment. The patient sample: overt and covert selection factors and...

(1) Our intake sample (n = 214) shows the well-known preponderance of upper-middle and middle class; however, in our case this selection is wholly independent from the influence of psychotherapists. Mean 'duration of problem' was over 10 years for 31% and from 5 to 10 years for another 18%. The intake sample deviated significantly in 'pathological' direction in every one of the 12 test scale means of the Freiburg Personality Inventory (FPI) from the population means. (2) Comparisons of the averse (i.e. to psychotherapy) subsample versus the affine group yielded cues that the former sample was a bit less ill and came from a less sophisticated background. The averse sample had a near normal arithmetic mean in 'verbal fluency' whereas its affine counter group was significantly higher. Therapy refusers (a subgroup of the averse) were less ill and had more often come 'on their own' than the 3 other subsamples. Therapy completers (a subgroup of the affine) were more often married and had a higher proportion of patients with a very long (10 years) 'duration of problem'. (3) Therapists' selection of candidates for psychotherapy operated through prognostic predictions for the intake sample. The prognoses of the CC predictors were significantly more optimistic than those of the PT prognosticians. In regard to psychological tests both CC and PT prognostically favoured 'manifest anxious' and 'inhibited' (socially anxious) patients. This communal nucleus is expanded for PT to include less sociable but not undercontrolled patients, whereas CC prognosticians favoured depressed, inferior feeling, but more verbally intelligent patients with less psychosomatic instability. Thus, except perhaps for the CC therapists preferring verbal intelligence, our prognosticians resisted the famous attraction of the YAVIS patient (Y = young, A = attractive, V = verbal, I = intelligent, S = successful) [Goldstein, 1977 p. 6]. (4) Lot assignment achieved a random distribution. There is a nonsignificant tendency for the immediate therapy sample to be a bit more disturbed than the waiting group. The patients filling out session questionnaires might be a bit more disturbed than those not asked to fill these in.