The usual textbook definition of learning is "change in behavior." If this definition is adopted as the search criterion for identifying pertinent items for review in the science education literature, several difficulties are immediately encountered. The conceptualization of learning in this way could be used to subsume studies in evaluation, for example in curriculum evaluation where presumably the curriculum serves as a means for some modification of behavior. Or studies in attitude change or any study where achievement is measured might be appropriately included. However, if an attempt is made to narrow the range by adopting a notion such as Ausubel's "Meaningful Verbal Learning" or Piaget's notion of learning, in the broad sense, as simply another term for intelligence, then very few studies in science education could be identified as within either domain. Thus, the immediate problem is deciding what to review.