Multidimensional Constrained Test Assembly

Abstract

Two mathematical programming approaches are presented for the assembly of ability tests from item pools calibrated under a multidimensional item response theory model. Item selection is based on Fisher information matrix. Several criteria can be used to optimize this matrix. In this article, the A-criterion and the D-criterion are applied. Empirical examples for a two-dimensional mathematics item pool illustrate the methods. Both criteria provided good results. Recommendations are provided about when to apply either approach.