How Binary Skills Obscure the Transition from Non-Mastery to Mastery

Autor/in	Karelitz, Tzur M.
Titel	How Binary Skills Obscure the Transition from Non-Mastery to Mastery
Quelle	In: Measurement: Interdisciplinary Research and Perspectives, 6 (2008) 4, S.268-272 (5 Seiten)Infoseite zur Zeitschrift PDF als Volltext Verfügbarkeit
Sprache	englisch
Dokumenttyp	gedruckt; online; Zeitschriftenaufsatz
ISSN	1536-6367
Schlagwörter	Stellungnahme; Feedback (Response); Item Response Theory; Predictor Variables; Educational Diagnosis; Diagnostic Tests; Psychometrics; Cognitive Measurement; Cognitive Processes; Cognitive Structures; Measurement Techniques; Evaluation Criteria; Evaluation Problems; Educational Assessment + Suchen Sie Ihr Suchwort? Item-Response-Theorie; Prädiktor; Pedagogical diagnostics; Pädagogische Diagnostik; Diagnostic test; Diagnostischer Test; Psychometry; Psychometrie; Cognitive process; Kognitiver Prozess; Cognitive structure; Kognitive Struktur; Messtechnik; Education; assessment; Bewertungssystem
Abstract	What is the nature of latent predictors that facilitate diagnostic classification? Rupp and Templin (this issue) suggest that these predictors should be multidimensional, categorical variables that can be combined in various ways. Diagnostic Classification Models (DCM) typically use multiple categorical predictors to classify respondents into qualitatively meaningful latent classes and, thus, provide a fine-grained analysis of respondents' strengths and weaknesses. The diagnostic power of DCM is driven by the confirmatory nature of the Q-matrix, which commonly represents item requirements in terms of a set of binary skills. Rupp and Templin note that, apart from a few exceptions, "Most DCM and associated estimation routines allow only for dichotomous latent variables." Binary skills are assumed to represent very simple dimensions in the sense that one either possesses a certain skill (mastery) or not (non-mastery). Such binary skills are useful indicators because it is easy to identify when something is present or absent. There is a slight problem, however. In reality, things are rarely black and white. In addition to being masters and nonmasters, individuals often fall into intermediate categories (e.g., partial masters). While models based on a continuous approach (e.g., Item Response Models) allow for infinite gradation along a proficiency continuum, the binary approach assumes no gradation in skill mastery. A person who was once a non-master (assumed to almost always fail items that require the relevant skill) could turn overnight into a master (assumed to almost always succeed on these items). For most practical purposes, it is necessary to be able to assess intermediate stages to allow for greater differentiation of mastery than what is made possible by the two extremes. Moreover, to be truly diagnostic, these intermediate stages need to be fully connected to relevant theory. In this commentary, therefore, the author suggests that relying on binary skills can obscure meaningful intermediate categories. Moving beyond such dichotomies would yield a more valid and realistic assessment of cognitive structures and processes; one that, ultimately, may be more useful to practitioners. (Contains 1 footnote. (ERIC).
Anmerkungen	Psychology Press. Available from: Taylor & Francis, Ltd. 325 Chestnut Street Suite 800, Philadelphia, PA 19106. Tel: 800-354-1420; Fax: 215-625-2940; Web site: http://www.tandf.co.uk/journals
Erfasst von	ERIC (Education Resources Information Center), Washington, DC
Update	2017/4/10