Countable indices of writing quality: Their suitability for screening‐eligibility decisions

Abstract

We examined five countable indices of writing quality for suitability in making special education screening‐eligibility decisions, collecting writing samples for 2,160 students in Grades 2 to 11 from two school districts. We examined the sensitivity of the writing indices around potential screening cutoff points on histograms and percentile line graphs with 68% standard error bands. In addition, we assessed criterion validity within each grade level by correlating the writing indices with teachers' holistic judgments of writing quality. Of the five indices, the percentage of words spelled correctly in a 3‐minute writing sample showed greatest measurement sensitivity for screening applications. If Grade 2 were excluded, the percentage of correct word sequences could be recommended as an alternative scoring method. Large standard errors of measurement, however, indicated the need for caution in interpreting score differences. Because we could not reliably differentiate scores 30 to 40 percentile points apart, grouping students from screening results may result in an unacceptable number of false‐positives and false‐negatives.