A Simplified Estimation of Latent State--Trait Parameters

Autor/inn/en	Hagemann, Dirk; Meyerhoff, David
Titel	A Simplified Estimation of Latent State--Trait Parameters
Quelle	In: Structural Equation Modeling: A Multidisciplinary Journal, 15 (2008) 4, S.627-650 (24 Seiten) PDF als Volltext Verfügbarkeit
Sprache	englisch
Dokumenttyp	gedruckt; online; Zeitschriftenaufsatz
ISSN	1070-5511
Schlagwörter	Structural Equation Models; Test Theory; Reliability; Sample Size; Monte Carlo Methods; Generalizability Theory + Suchen Sie Ihr Suchwort? Testtheorie; Reliabilität; Monte-Carlo-Methode
Abstract	The latent state-trait (LST) theory is an extension of the classical test theory that allows one to decompose a test score into a true trait, a true state residual, and an error component. For practical applications, the variances of these latent variables may be estimated with standard methods of structural equation modeling (SEM). These estimates allow one to decompose the coefficient of reliability into a coefficient of consistency (indicating true effects of the person) plus a coefficient of occasion specificity (indicating true effects of the situation and the person-situation interaction). One disadvantage of this approach is that the standard SEM analysis requires large sample sizes. This article aims to overcome this disadvantage by presenting a simple method that allows one to estimate the LST parameters algebraically from the observed covariance matrix. A Monte Carlo simulation suggests that the proposed method may be superior to the standard SEM analysis in small samples. (Contains 1 figure and 3 tables.) (As Provided).
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