Cognitive utilization of probabilistic cues

This article is concerned with a laboratory experiment on a classificatory process similar to medical diagnosis. S s were required to learn the names and descriptions of 16 objects where the cues comprising these descriptions were probabilistic in nature. They then solved a number of object identification problems consisting of the presentation of a single cue, followed by a sequence of yes‐no questions about the presence or absence of other cues, and terminating with the identification of the object. The results were interpreted as supporting a serial hypothesis‐testing model and reveal how such hypotheses are generated and confirmed. The study is viewed as a necessary step toward the creation of a computer model of medical diagnosis. The model is briefly described.