Analogy with knowledgeable learners: when analogy confers benefits and exacts costs.

Though embedding analogy in a text can improve acquisition of concepts that are relatively unfamiliar to learners (Donnelly & McDaniel, 1993), it remains uncertain how analogy influences learning for learners with some background knowledge for the target material. We develop several theoretical possibilities and report an experiment to test these possibilities. Target concepts familiar to college students were expressed either in literal form or through analogy, and the concepts were presented in a manner that either facilitated or discouraged contact with subjects' prior knowledge. As revealed by multiple-choice testing, analogy facilitated learning when the targets were presented to discourage contact with prior knowledge. In contrast, analogy decreased performance when the same targets were presented to facilitate contact with prior knowledge. Possible interpretations of this unique finding are discussed.