The Decision-making Utility and Predictive Power of DIBELS for Students' Reading Achievement in Pennsylvania's Reading First Schools
Year: 2007
Type: article
Study design: Observational Study
Abstract: The purpose of this study was to examine the predictive strength and decision-making utility of the Dynamic Indicators of Early Literacy Skills (DIBELS). Specifically, the study examined whether DIBELS benchmarks correctly differentiated among students who were at-risk for reading failure and those who were not as measured by end-of-third-grade achievement on the Pennsylvania System of School Assessment (PSSA) in Reading First schools. More broadly, this study addressed the effectiveness of DIBELS for early identification of children considered to be at-risk for reading failure using the author-recommended benchmarks. Additionally, data were analyzed to determine whether first grade cut-points were appropriately sensitive and specific in relation to long-term predictions (end of third grade) of special education status. When comparing within-year achievement trends, results indicated that DIBELS was generally predictive of first through third grade students' Fall to Spring achievement. However, some students did demonstrate erratic achievement. Receiver Operating Characteristic (ROC) analyses revealed that the author-recommended cut-points for the Fall subtests resulted in concerning numbers of false negative and false positive predictions of reading achievement. In fact, the cut-points for the phoneme segmentation fluency (PSF) subtest were found to have a statistically inappropriate balance of sensitivity and specificity. Hierarchical Linear Modeling (HLM) analyses of students' long-term achievement showed that the DIBELS measures administered early in first grade were generally not predictive of third grade reading achievement for students in these Reading First schools. In fact, first grade results explained only 18% of the variability in students' third grade reading scores on the PSSA. Finally, logistic regression results suggest that students' socio-economic status and race were more accurate predictors of end-of-third grade special education status than their first grade reading achievement on the DIBELS. The overall limited predictive value of DIBELS on students' long-term reading achievement raises important concerns about over-reliance on DIBELS in an early intervention framework like Pennsylvania's Reading First initiative and in school-wide educational decision making systems such as Response-to-Intervention (RTI).
Author: Amanda Kloo
Language: English
Subfield: Developmental and Educational Psychology
Field: Psychology
Domain: Social Sciences
SDG: Quality education
Open Access status: green
OpenAlex ID: W2132785789