Using Math and Reading Curriculum-Based Measurements to Predict State Mathematics Test Performance

Abstract

This study investigated the technical adequacy of 3 simple, 1-min curriculum-based measurements (CBM) for predicting third (n = 35) and fifth (n = 49) graders' performance on a state standards test of mathematics. Measures included 2 numeric CBMs: a traditional basic facts measure and a new “cloze” math facts measure, designed to require deeper understanding of fact families. A silent reading measure served as third predictor CBM. Results indicated that reliability and criterion validity of 1-min administrations were insufficient. When scores from 2 administrations were aggregated, moderate correlations were achieved. Results of aggression analysis showed that reading and cloze math measures each made significant unique contributions to explaining performance on the state math test and together explained 52% of variance at fifth grade and 27% at third grade.