Evaluation of Structure Complexity Magnitude, Degree of Cross-Loading on Secondary Dimension and Model Specification on MIRT Parameter Estimation

Autor/in	Hosseinzadeh, Mostafa
Titel	Evaluation of Structure Complexity Magnitude, Degree of Cross-Loading on Secondary Dimension and Model Specification on MIRT Parameter Estimation
Quelle	(2021), (133 Seiten) PDF als Volltext Verfügbarkeit Ph.D. Dissertation, Oklahoma State University
Sprache	englisch
Dokumenttyp	gedruckt; online; Monographie
ISBN	979-8-4387-5039-0
Schlagwörter	Hochschulschrift; Dissertation; Item Response Theory; Models; Simulation; Evaluation Methods; Maximum Likelihood Statistics; Mathematics; Item Analysis; Correlation; Data; Error of Measurement; Statistical Bias + Suchen Sie Ihr Suchwort? Thesis; Dissertations; Academic thesis; Item-Response-Theorie; Analogiemodell; Simulation program; Simulationsprogramm; Mathematik; Itemanalyse; Korrelation; Daten; Messfehler
Abstract	In real-world situations, multidimensional data may appear on large-scale tests or attitudinal surveys. A simple structure, multidimensional model may be used to evaluate the items, ignoring the cross-loading of some items on the secondary dimension. The purpose of this study was to investigate the influence of structure complexity magnitude of the data incorporating the degree of cross-loading on secondary dimensions and model specification, especially when the model was misspecified as a simple structure, ignoring the cross-loading, while the data are truly complex on item parameter recovery in MIRT models. In order to address the research question a simulation study that replicated this scenario was designed in order to manipulate the variables that could potentially influence the precision of item parameter estimation in the MIRT models. Item parameters were estimated using marginal maximum likelihood (MML), utilizing the expectation-maximization (EM) algorithms. A compensatory 2PL-MIRT model with two dimensions and dichotomous item response type (Reckase, 1985) was used to simulate and calibrate the data for each combination of conditions across 500 replications. The result of this study indicated that ignoring complex structure of the multidimensional data incorporating the degree of cross-loading and model specification severely impact item discrimination estimations resulting biased and inaccurate item discrimination parameters. When the complex structure of the data was misspecified, whether the data were correlated or uncorrelated, item discrimination parameters were adversely affected. As the complexity magnitude incorporating the degree of cross-loading increased, the error and bias estimates of item discrimination worsened. Furthermore, the results of this study indicated that if the data are correlated and the correlation is not specified nor are the item cross-loadings, item discrimination estimates, specifically for the truly cross-loaded items had extremely poor error and bias estimates. [The dissertation citations contained here are published with the permission of ProQuest LLC. Further reproduction is prohibited without permission. Copies of dissertations may be obtained by Telephone (800) 1-800-521-0600. Web page: http://www.proquest.com/en-US/products/dissertations/individuals.shtml.] (As Provided).
Anmerkungen	ProQuest LLC. 789 East Eisenhower Parkway, P.O. Box 1346, Ann Arbor, MI 48106. Tel: 800-521-0600; Web site: http://www.proquest.com/en-US/products/dissertations/individuals.shtml
Erfasst von	ERIC (Education Resources Information Center), Washington, DC
Update	2024/1/01