An Item Fit Statistic Based on Pseudocounts from the Generalized Graded Unfolding Model: A Preliminary Report.Reportar como inadecuado




An Item Fit Statistic Based on Pseudocounts from the Generalized Graded Unfolding Model: A Preliminary Report. - Descarga este documento en PDF. Documentación en PDF para descargar gratis. Disponible también para leer online.





Stone and colleagues (C. Stone, R. Ankenman, S. Lane, and M. Liu, 1993; C. Stone, R. Mislevy and J. Mazzeo, 1994; C. Stone, 2000) have proposed a fit index that explicitly accounts for the measurement error inherent in an estimated theta value, here called chi squared superscript 2, subscript i*. The elements of this statistic are natural byproducts of the marginal maximum likelihood procedure used to estimate generalized grading unfolding model (GGUM) parameters. The objective of this study was to generalize the proposed fit index to the GGUM situation. The behavior of the statistic under the GGUM was examined using alternative simulation techniques, and the results were used to develop hypothesis testing criteria for item misfit assessment. The number of simulated questionnaire items and the number of hypothetical respondents were both varied using a 3 x 4 factorial design, and the number of questionnaire items was 10, 20, or 30, with 500, 750, 1000, or 2000 simulees. Preliminary results suggest that the chi squared superscript 2, subscript i* family of item fit assessment methods is promising in the context of the GGUM. (Contains 1 table, 2 figures, and 21 references.) (SLD)

Descriptors: Chi Square, Goodness of Fit, Test Items











Autor: Roberts, James S.

Fuente: https://eric.ed.gov/?q=a&ft=on&ff1=dtySince_1992&pg=6166&id=ED476920







Documentos relacionados