This study investigated the factorial structure of changes concomitant with client-centered therapy. Most of the variables were differences obtained by subtracting the pretherapy scores from the corresponding posttherapy scores of each of 93 clients. The 1st 4 factors seem to reflect: change in favorableness of client self-evaluation, change in adequacy ratings based on the TAT, therapist's perception of change, and change on Hs and Hy from the MMPI. The 5th factor was length of therapy. Of these orthogonal factors, the 1st 4 are seen as method factors, each being associated primarily with an observer-instrument combination. As measured, the changes associated with short psychotherapy cannot be adequately represented by a single global rating or by scores for personality traits denned either broadly or narrowly.

An attempt to predict change in psychotherapy. Criteria included composite raw and residual gain scores based on factor analyses. Predictors included pretherapy scores on the criterion dimensions and on other variables. Intercorrelations within each set of criteria were generally low although some were significant. The few significant correlations between predictors and criteria did not reveal any systematic pattern except for pairs of variables derived from the same rater or shared sources of data. For clients at a university counseling center, raw or residual gains over relatively short psychotherapy are so specific to the given method that they are difficult or impossible to predict from measures based on independent methods.

No abstract

Cartwright (5) has presented evidence for the existence of a "failure-zone," ranging between 13 and 21 interviews, in client-centered therapy. In a subsequent paper, Taylor (25) presented closely comparable evidence with respect to psychoanalytically oriented psychotherapy.Cartwright demonstrated that therapeutic outcomes were divided by the failure-zone and suggested that there may be two forms of therapeutic process, identifying these processes as "short" (1-12 interviews) and "long" (13-77 or more interviews). He further hypothesized (5, p. 362) "that certain individual differences between clients give rise to different kinds of therapeutic process."On the basis of these results and hypotheses, five client groups were distinguished: short success, 1-12 interviews; short failure, 1-12 interviews; failure-zone, 13-21 interviews; long success, more than 21 interviews; and long failure, more than 21 interviews. Utilizing this schema, Kirtner (12) hypothesized that therapy length-by-outcome is related to the personality structures of clients at the beginning of therapy. The present paper reports an investigation of this hypothesis. Method and Procedure SubjectsFrom the research files of the Counseling Center, University of Chicago, all clients were selected for whom the following four condi-1 This investigation is based upon a dissertation submitted by the senior author to the University of Chicago in partial fulfillment of the requirements for the degree of master of arts.

REFERENCES 1. CARTWRIQHT, D. S. Succesa in psychotherapy aa a function of certain acturial variables.