CUES, COMPONENTS, AND CAUSAL STRUCTURE IN LABORATORY JUDGMENT STUDIES

digms generally ignore the issue of the causal structure underlying the correlations between the cue(s) and the distal variable. The present note argues that two possible causal structures must be considered: cue-type tasks, in which the stimuli reflect the value of the distal variable, and component-type tasks, in which the stimuli values cause the distal variable value. It is shown that different optimal response strategies can be derived for a given set of cue validities, depending on which causality assumptions are made, and these optimal strategies are derived for the simple cases for n = 1, 2, and 3. Some implications of these results for the design and inter-pretation of laboratory studies are suggested. THE intent of this note is to propose a distinction between two classes of problems used in laboratory studies of judgment and infer-ence phenomena. The distinction is relevant to both single-cue proba-bility learning (SCPL) and multiple-cue probability learning (MCPL) tasks within the linear lens model paradigm. Some implications of the distinction for the conduct and interpretation of such laboratory stud-ies are noted. Cues versus Components The usual laboratory procedure in MCPL studies consists of a series of trials in each of which the subject is shown a stimulus vector (Xl, X2