Transfer of schema learning to schematic concept formation

Abstract

Schema theorists have proposed the necessity for humans to retain the schematic aspects of experience. As a mnemonic mechanism, the concept of schemata has been hypothesized. Recent research in the context of schema theory has benefited greatly by the development of procedures whereby experimentally controlled stimulus populations can be generated so as to exhibit multivariate schematic attributes. Ability to encode the prototypical or schematic characteristics of two or more different schema populations, simultaneously present in the environment, on the basis of information derived from perceiving sampled instances has been termed schematic concept formation (SCF). This study tested the important assumption of schema theory that once a schema is learned or encoded it continues to be available for use in other situations. The purpose of this study was to demonstrate that prior schema encoding would facilitate subsequent SCF when the previously encoded schema was one of two schema populations in the SCF situation. Three experiments were conducted employing the specific transfer paradigm. Independent groups of Ss received pretraining with a single schema that either was or was not the same as one of two schemata to be differentiated in a SCF transfer task. The first two experiments were essentially the same in that the single schema pretraining task required similarity judgments to be made on randomly sampled pairs of schema instances. The SCF task was similar, but stimulus pairs were randomly sampled from two schema populations. Subjects in the third experiment were asked to categorize individually presented schema instances in both the single schema pretraining and two-schemata SCF transfer tasks. Only the third experiment constituted an effective demonstration of single schema transfer to SCF. The conclusion was that the single stimulus presentation was the critical factor in establishing a true memory requirement on Ss.