Different approaches to equating oral reading fluency passages

Abstract

Using curriculum-based measures (CBM) to identify and monitor students’ oral reading fluency (ORF) is challenging, with student performance subject to numerous sources of variability. One source of variability that is beyond teachers’ or students’ control stems from differences in text difficulty across CBM probes at any given grade level. These differences are referred to collectively as form effects on students’ ORF. This chapter examines the research on form effects and different solutions that have been discussed in the research literature for reducing or removing form effects from CBM assessments. These solutions are referred to collectively as equating methods. The chapter examines four different equating methods using data from a sample of 1867 students from grade 6-8 who were evaluated on subtests of the Texas Middle School Fluency Assessment (TMSFA) to illustrate the differences across the methods. These methods either focus on the equating of raw scores, or on the estimation of true fluency scores through the modeling of test forms. The raw score methods include linear and equipercentile equating, while the true score methods include linear and nonlinear equating using latent variables (LVs). The results are discussed in terms of their implications for developers and users of CBM assessments.