Using Item Response Theory to Assess Effects of Mathematics Instruction in Special Populations

Abstract

Research on mathematics education often includes some experimental manipulation of instruction in order to compare the effects of one type of instruction with that of one or more others. Tracking students' progress during treatments may be difficult in studies with special populations where numbers of participants are sometimes very small. In this study, we illustrate a method for scaling and reporting test results that helps to overcome this problem, particularly when number-correct scoring is used on measures to assess treatment effectiveness. Results are compared for an instructional experiment with remedial and nonremedial students using a number-correct scoring and a standardized score scale developed using item response theory (IRT).