Development and Application of a Model of Individual Decision Making

摘要

The major purposes of our grant were: (a) to develop a psychological model ofhuman reasoning in situations in which people are told some facts and then have to assign probabilities to generalizations of these facts; and (b) to use this model to develop reasoning prosthetics, i.e., automated procedures that enable an individual reasoner to increase the accuracy of his or her probability judgments while remaining faithful to one's core intuitions. With respect to our first goal, we developed and tested the Gap model, which assumes that people's probability judgements about a domain are mainly based on their perceived similarity relations among objects in the domain, and on their intuitions about the plausibility of object-predicate relations in the domain. Quantitative versions of the Gap model were shown to do an adequate job of predicting people's probability judgments. With respect to our second goal, we have developed a number of reasoning prosthetics, each of which can reliability predict people's actual probability judgments, and in some cases more accurately match the true probabilities than can people's judgments. Other topics in reasoning and probability judgments were also studied.