Theories regarding the impact of therapeutic styles on patient outcomes are reviewed, followed by a description of the methodology used by John Whitehorn and Barbara Betz in their studies of psychotherapy effective with schizophrenia. Using records from the Phipps Clinic at Johns Hopkins University Medical School, Baltimore, Md., these researchers developed a scale of 23 items from the Strong Vocational Interest Bank (SVIB) that differentiated 'A' doctors who had high rates of success with schizophrenics from 'B' doctors who had low rates. A preliminary study which measured the personalities of 60 male college students who had rated pictures drawn by mental patients in another experiment included items from the A-B scale and demonstrated that the A-B predictor was limited in nonhospital settings and might even reverse its meaning. The A-B methodology was then extended to a larger sample of 142 mental health workers. Participants completed a Personal Tendencies Questionnaire (PTQ) and the SVIB. The resulting data was correlated with respondents' percent improvement in treating patients who were schizophrenics, depressives, and neurotics. A separate analysis was conducted for doctors who prescribed drugs for their schizophrenic patients. A third study focused on 51 helpers -- counselors and house parents -- at the Loysville Youth Development Center, a diagnostic facility part of Pennsylvania's juvenile justice system, in an effort to examine the characteristics of therapists who were effective with personality disorders in a milieu setting. Helpers completed the PTQ, SVIB, and the Quay Correctional Preference Test, and their responses were correlated with improvement scores for neurotics and five other diagnostic groups. Findings from all of these studies suggest that neurotics are not well served by therapists who like to solve problems, while schizophrenics are best helped by a therapist who is active and personally involved with the patient. This research supports the theory that different disorders require different treatments. Tables, over 125 references, and indexes by name and subject are provided. The appendixes contain detailed descriptions of statistical methods and the A-B scale, the PTQ, supplemental analyses of data used in the studies, Pennsylvania's diagnostic classification system, and additional information on drugs and the therapeutic relationship.