Configural judgments revealed (summary)

摘要: Studies of clinical judgment can be divided into two groups; those in the first group focus upon outcome; those in the second, upon the judgment process itself, particularly the manner in which cues are weighted and combined by a judge. This study, which demonstrates the use of structural models, such as those underlying the analysis of variance, for the representation of clinical judgments, falls in the second group. METHOD Six physicians (P), six nurses (N), six nurse's aides (A), three psychologists (), and three social workers (S) from three psychiatric wards of a VA hospital decided, independently, on the basis of written descriptions, whether they would or would not grant a weekend pass to each of 128 presumably real, but actually hypothetical, patients. They also indicated their confidence in each judgment with one of six numbers:. 5,. 6,. 7,. 8,. 9, 1.0. These confidence judgments were converted to a single ordinal scale, presumably representing the probability of, or the confidence in, granting a pass to the patient for the weekend. The patient descriptions were constructed by indicating Yes or No for each of six questions, as illustrated by the description of the first patient in Table 1. On the basis of considerable preliminary study and discussion, it was concluded that these six items probably