Detecting Cognitive Change in the Math Skills of Low-Achieving Adolescents

Abstract

Current methods for detecting growth of students’ problem-solving skills in math focus mainly on analyzing changes in test scores. Score-level analysis, however, may fail to reflect subtle changes that might be evident at the item level. This article demonstrates a method for studying item-level changes using data from a multiwave experiment with a teaching method called enhanced anchored instruction (EAI). The analysis combines a mixture Rasch model for detecting individual differences within latent groups with a latent transition analysis model for tracking changes in latent group membership over the course of EAI. The analysis clearly indicates the effects of EAI and how they differ for members of each latent class. Comparisons are provided with a standard analysis of changes in test scores. Implications of the new approach are discussed for detecting subtle transformations in the math performance of students across a range of ability levels.