A HIERARCHICAL ITEM RESPONSE MODEL FOR EDUCATIONAL TESTING

摘要: Publisher Summary This chapter presents the item–response model and its use for educational testing. In a review of IRT methodology assessment, Bock, Mislevy, Woodson outline two approaches well-suited to population focus assessment. The first approach is more familiar person-level models, though bypassing computation results by estimating item parameters directly from counts response patterns. second group-level models in narrowly-defined content areas. Both have disadvantages. duplex logical extension if information about individuals broadly-defined area also desired. It maintains model's advantage maximum efficiency results, imposes less computational burden than approach. achieved at cost restrictive distributional assumptions, such as homoscedasticity within groups over time. Rather characteristics finite populations that constitute, presented explains performance manifestation processes under control latent structure, estimates characterize structure.