An Exploration of Alternative Scoring Methods Using Curriculum-Based Measurement in Early Writing

Abstract

This manuscript describes two empirical studies of alternative scoring procedures used with curriculum-based measurement in writing (CBM-W). Study 1 explored the technical adequacy of a trait-based rubric in first grade. Study 2 explored the technical adequacy of a trait-based rubric, production-dependent, and production-independent scores in third grade. Results of Study 1 suggest that the rubric holds promise as a valid measure of sentence writing ability in first grade and has utility as a supplemental scoring procedure when using CBM-W as a screening tool. Results of Study 2 show that correct word sequences maintained the highest correlation coefficients across time with the trait-based rubric, but the other scoring procedures might offer promise as reliable alternative scoring methods. However, high internal correlations among the text features of the rubric along with highly variable interrater reliability suggest that caution must be taken in interpreting results.