The Development of Automaticity in Short-Term Memory Search: Item-Response Learning and Category Learning

Autor/inn/en	Cao, Rui; Nosofsky, Robert M.; Shiffrin, Richard M.
Titel	The Development of Automaticity in Short-Term Memory Search: Item-Response Learning and Category Learning
Quelle	In: Journal of Experimental Psychology: Learning, Memory, and Cognition, 43 (2017) 5, S.669-679 (11 Seiten) PDF als Volltext Verfügbarkeit
Sprache	englisch
Dokumenttyp	gedruckt; online; Zeitschriftenaufsatz
ISSN	0278-7393
DOI	10.1037/xlm0000355
Schlagwörter	Short Term Memory; Recall (Psychology); Item Response Theory; Learning Processes; Cognitive Mapping; Experiments; Classification; Recognition (Psychology); Efficiency; Undergraduate Students; Visual Stimuli; Indiana + Suchen Sie Ihr Suchwort? Kurzzeitgedächtnis; Abberufung; Item-Response-Theorie; Learning process; Lernprozess; Erprobung; Classification system; Klassifikation; Klassifikationssystem; Recognition; Wiedererkennen; Effectiveness; Effektivität; Wirkungsgrad
Abstract	In short-term-memory (STM)-search tasks, observers judge whether a test probe was present in a short list of study items. Here we investigated the long-term learning mechanisms that lead to the highly efficient STM-search performance observed under conditions of consistent-mapping (CM) training, in which targets and foils never switch roles across trials. In item-response learning, subjects learn long-term mappings between individual items and target versus foil responses. In category learning, subjects learn high-level codes corresponding to separate sets of items and learn to attach old versus new responses to these category codes. To distinguish between these 2 forms of learning, we tested subjects in categorized varied mapping (CV) conditions: There were 2 distinct categories of items, but the assignment of categories to target versus foil responses varied across trials. In cases involving arbitrary categories, CV performance closely resembled standard varied-mapping performance without categories and departed dramatically from CM performance, supporting the item-response-learning hypothesis. In cases involving prelearned categories, CV performance resembled CM performance, as long as there was sufficient practice or steps taken to reduce trial-to-trial category-switching costs. This pattern of results supports the category-coding hypothesis for sufficiently well-learned categories. Thus, item-response learning occurs rapidly and is used early in CM training; category learning is much slower but is eventually adopted and is used to increase the efficiency of search beyond that available from item-response learning. (As Provided).
Anmerkungen	American Psychological Association. Journals Department, 750 First Street NE, Washington, DC 20002. Tel: 800-374-2721; Tel: 202-336-5510; Fax: 202-336-5502; e-mail: order@apa.org; Web site: http://www.apa.org
Erfasst von	ERIC (Education Resources Information Center), Washington, DC
Update	2020/1/01