Master's Theses



Degree Name

Master of Science (MS)


A multidimensional scaling method was applied to nonclient preferences about psychotherapy as a beginning step toward establishing the basic dimensions along which therapy preferences may be described. Fifty-one therapy preferences statements were generated on the basis of a conceptual framework involving a miscellaneous category and Rosen’s (1967) five a priori categories (Counselor Attributes; Counselor Personality; Counseling Procedures; Problems and Topics; Counseling Resources). These preferences were then presented to 84 nonclient university students in the form of a card deck, with one preference typed on each card. Participants were asked to sort the 51 preferences into five categories on the basis of similarity. The number of times any two preferences were placed together was recorded and entered into a data matrix. The matrix was then analyzed by the ALSCAL program in the SAS computer system. A four-dimensional solution appeared to provide the most satisfactory representation of the data. Dimension 1 seemed to represent the preferences in terms of what type of help is preferred: preference for a non-psychotherapeutic, medical type of help seemed reflected at the positive end of this dimension while preferences for a psychotherapeutic relationship seemed reflected by the negative end. The positive and negative extremes of Dimension 2 appeared to represent preferences about therapist attributes and therapy content (problems and topics), respectively. The third dimensions seemed to have one uninterpretable end while the other end seemed exemplary of preferences with regard to therapy resources. Dimensions 4 also appeared to have an uninterpretable extreme, with the other extreme seeming to concern preferences about therapy procedures. These findings were discussed in terms of the adequacy of Rosen’s categories as a conception of the domain of therapy preferences. While the results appeared to confirm four of the categories, none of the categories seemed to clearly reflect the non-psychotherapeutic extreme of Dimension 1. Also, the Counselor Personality category appeared to be only partly and obliquely reflected in the psychotherapeutic extreme of Dimension 1 (and to a lesser extent in the extremes of any other dimension). It was suggested that whether Rosen’s categories need refinement could best be determined by further studies to investigate the reliability of Dimension 1. These studies would need to include items that made a clear distinction between preference for help afforded by an interpersonal relationship and preference for help afforded by the application of a relatively impersonal procedure (e.g., course of medication, relaxation exercises). Finally, the present findings were viewed as suggesting that it may be closer to an individual’s experience to construe therapy preferences as unipolar dimensions (involving greater or lesser quantities of a characteristic) rather than as true, bipolar dimensions.


Richard P. Schellenberg

Date of Award

Fall 1981

Document Type

Thesis - campus only access


© 1981 Bob A. Campbell


For questions contact

Off Campus FHSU Users Click Here