It is a common phenomenon of clinical practice that as a psychologist accumulates case-study material about another human being, he comes to think that he knows that person pretty well. Consequently, sooner or later in the information-gathering process, the psychologist becomes confident enough to make diagnostic conclusions, describe the client's main dynamics, and perhaps even venture to predict his future behavior. Though the psychologist's conclusions may remain tentative, his increase in confidence from the time of first approaching the case to the time of writing his report is usually very marked.

This study investigated whether that increase in confidence is justified by a corresponding increase in accuracy of conclusions. Though the psychologist's confidence in his conclusions has often been mentioned as an important subject of scientific inquiry (Meehl, 1957), it has only rarely been studied intensively. Furthermore, when it has been studied, rather surprising findings have often resulted. For instance, L. R. Goldberg (1959) and Oskamp (1962) have shown that the diagnostic confidence of experienced psychologists is less than that of less experienced persons. The same studies and many others have also shown that professional psychologists are no better interpersonal judges, and sometimes are worse ones, than are untrained individuals (Taft, 1955).

Another rarely studied factor, which may provide a good index of the expertness of a judge, is the relationship between his level of confidence and his level of accuracy.