The Rorschach test in clinical diagnosis--a brief commentary.

I am pleased to have this opportunity to respond to the article on the Rorschach Test in clinical diagnosis since my 1947 article seemed to be at least a partial focus of the critical comments. I also will try to comply with the editor’s request for a “short rejoinder” to this lengthy article. In addition, although I felt the tone of the manuscript was overly haughty in criticizing an article published 52 years ago, I will attempt an objective response. First, it is important to remember that the Rorschach was a relatively new test in 1946 when my report was prepared, and few research studies had been conducted. There were just a small number of clinical-psychology training programs in 1948, and the Division of Clinical Psychology had 482 members. The emphasis on the Rorschach was on personality appraisal and clinical diagnosis rather than on psychometric scores. It was the new era of projective techniques as contracted with psychometric questionnaires. After praising me for my contributions to clinical psychology, the authors state: “It is instructive, therefore, to see how this careful and thoughtful scholar was led to conclusions regarding the Rorschach that were almost certainly in error . . .” (p. 396). They then discuss three significant errors that they state can still be found in current Rorschach research. The first refers to the “set of dramatic results” that appear to support the value of the Rorschach for diagnostic purposes. The authors performed their calculations on the figures provided and have come up with a .90 correlation with staff diagnoses of schizophrenia where I stated that 71.9% of the Rorschach diagnoses were in agreement with the staff diagnoses. Whereas, it is stated, such high “validity coefficients” would arouse