ABSTRACT
Good readers can recognize printed words more quickly than can poor readers (Perfetti, 1985). Chess experts can remember game positions more accurately than novices (Chase & Simon, 1973). Skills such as reading recognition and chess memory are emergent properties stemming from well-developed, easily accessible knowledge bases built from extensive reading or chess experience. These skills also can serve as limited proxy measures for assessing reading and chess ability, particularly if one wants to make a crude cut between good and poor performers. That is not to say, however, that such proxy measures are adequate for all assessment decisions, particularly those associated with instruction, for at least two reasons. First, these findings do not necessarily imply that training the emergent properties will benefit the criterion performance. Teaching someone skill at remembering chess positions is not likely to improve their overall chess playing ability. Second, because these proxy measures are not tied to the epistemology of a domain, the relationship of the skill to the domain is murky. Such tenuous relationships frequently lead to misleading instructional advice.
