The Predictive Utility of Kindergarten Screening for Math Difficulty

This study examined the reliability, validity, and predictive utility of kindergarten screening for risk for math difficulty (MD). Three screening measures, administered in September and May of kindergarten to 196 students, assessed number sense and computational fluency. Conceptual and procedural outcomes were measured at end of first grade, with MD operationalized as below the 16th percentile. The authors compared single-versus multiple-skill screeners, fall versus spring kindergarten screening, and conceptual versus procedural outcomes. Reliability and validity coefficients were adequate. Logistic regression and receiver operating characteristics analyses indicated that the single- and multiple-skill screeners produced good and similar classification accuracy at the fall and spring screening occasions in forecasting conceptual outcome. To forecast procedural outcome, the screeners produced similar but less accurate fits.