engleski

Comparison of Rasch Model and One Parameter Logistic Model on Croatian National Assessment Data

In the international project “Development of Instruments in Croatian National Assessment” Croatian National Centre for External Evaluation of Education (NCEEE) cooperated with Institute for Educational Measurement, Netherlands (CITO). In analysis of Mathematics data One Parameter Logistic Model (OPLM) was applied. OPLM is a model where difficulty parameters are estimated and discrimination indices are imputed as known constants, which is an extension of the Rasch model. The aim of this paper is to compare two different IRT models: Rasch model and OPLM in the terms of fitting statistics and ability distributions. The goodness-of-fit of an item set with the OPLM is investigated with five goodness- of-fit statistics. Results of 1229 gymnasium students on three Mathematics domains were used for the purpose of this research. OPLM software were used in the analysis of results (Verhelst, Glas & Verstralen, 2005) was used. Estimation through Conditional Maximum Likelihood method showed superiority of OPLM in compare to Rasch model for the data that were used. When using OPLM, it is possible to find a model that will have better fitting statistics than Rasch model. OPLM combines attractive mathematical properties of the Rasch model with the flexibility of the two-parameter model.

Rasch model; One Parameter Logistic model; comparison

nije evidentirano