THE ABSTRACTION OF INTERVENING CONCEPTS FROM EXPERIENCE WITH MULTIPLEINPUT-MULTIPLE OUTPUT CAUSAL ENVIRONMENTS

The purpose of this article is threefold: (a) introduce a new paradigm for investigating how intervening concepts are learned, (b) report fo ur new experiments that provide converging evidence for the acquisitio n of intervening concepts, and (c) propose a simple associative learni ng mechanism to account for the results. The new paradigm utilizes a s timulus-response-feedback task in which subjects learn trial by trial how a multivariate set of inputs maps into a multivariate set of outpu ts. The first two experiments use evidence based on a principal compon ent analysis to replicate the finding that intervening-concept learnin g occurs spontaneously, but only in environments that contain an inter vening factor. The next experiment provides a second converging line o f evidence for this conclusion by showing that subjects can use an int ervening concept to make accurate inferences to a new fourth output du ring a transfer test. The last experiment provides a third line of evi dence by showing that subjects can use an intervening concept to make accurate inferences from a new fourth input. The results are explained by a hidden-unit connectionist learning mechanism that includes both accuracy and parsimony as learning objectives. (C) 1997 Academic Press .