Predicting Student Responsiveness to Fast ForWord Using DIBELS Subtests

Abstract

The current study was completed through a retrospective analysis of school records of elementary school students in the Northeast Region of the Philadelphia School District (PSD) who have participated in the Fast ForWord (FFW) Language program. The data requested from student records included: demographic information (e.g., gender, grade,age, ethnicity, disability, and special education status), DIBELS scores, and FFW completion and participation variables. The current study set out to determine if DIBELS scores can predict student performance or mastery level on the FFW program. A total of seven individual FFW variables (percent of completion for each activity) and five overall FFW variables (percent complete, participation level, attendance level, total days to complete, and successful performance) served as the outcome variables. Frequency distributions, Pearson correlations, an ANOVA, and a standard multiple regression were used to determine the relationships of demographic variables among predictor and outcome variables as well as the predictive power of the DIBELS test scores. Results of the standard multiple regression analysis failed to yield significant results in the ability for either DIBELS raw or benchmark scores to predict performance on the FFW reading program. The current study highlighted that in the real-life conditions of a large, urban public school system, DIBELS seems to have very little, if any, predictive abilities specific to designating students appropriately to an intensive, costly, and time-consuming intervention program. It is hoped that the information presented in this study will stimulate some positive discussion and changes in the assessment and referral processes currently being widely employed across American schools in order to better serve and educate American children that demonstrate symptoms of early reading deficits.