This report summarizes fifteen months of research on the technology of inference and decision. Theoretical and experimental work on two major topics; assessment of subjective probabilities and multiattribute utility theory, is discussed. Experimental work showed that individuals bringing orthogonal information to a group probabilistic inference task tend to average that information rather than multiplying as the normative rule indicates. Other experimental work indicated that for both continuous and descrete variables simple averaging of individual's probability judgments to form a group judgment did not differ significantly from behavioral interaction in final quality of the judgments as evaluated by a quadratic scoring rule. Several experiments in the area of assessing small probabilities serve to show that judgments of these small numbers are of strickingly low quality. Use of both the marker events technique and several computer program aids have not substantially improved these judgments.
Stored only when both papers are in the corpus — 1 citations exist upstream.