Some Posterior Standard Deviations in Item Response Theory.

A study was conducted to examine some posterior standard deviations in item response theory (IRT). Data were obtained primarily by using two sets of priors for different IRT models for sample sizes of 100 and 1,000. Findings revealed that use of priors may lessen the instability of estimates of it...

Descripción completa

Detalles Bibliográficos
Publicado en:Educational & Psychological Measurement Vol. 67; no. 2; pp. 258 - 280
Autor principal: Kim, Seock-Ho
Formato: Artículo
Publicado: Sage Publications Inc. April 2007
Materias:
Acceso en línea:Ver este registro en EBSCOhost
Descripción
Sumario:A study was conducted to examine some posterior standard deviations in item response theory (IRT). Data were obtained primarily by using two sets of priors for different IRT models for sample sizes of 100 and 1,000. Findings revealed that use of priors may lessen the instability of estimates of item parameters, given that the choice of priors is reasonable. Findings also showed that the sample size needed for acceptable accuracy for the practical application of item response theory can be inferred from computer programs and tables. Findings are discussed in detail.