Using Error Recognition to Identify Students at Risk of Reading Failure
The current trend in education points toward preventative methods in order to address the problem of reading failure. An essential component of a preventative method is early identification of students at risk. Effective screeners are those with high classification accuracy, efficiency, and usability. Though current screeners such as the Dynamic Indicators of Basic Early Literacy Skills (DIBELS) have sufficient classification accuracy for their intended purpose, there is room for improvement in the accuracy and usability of these measures. The forthcoming study will identify differences in the accuracy of an experimental error recognition measure, DIBELS Oral Reading Fluency (DORF) measure, and DIBELS Daze measure for identifying 3rd grade students at risk for reading failure. Error recognition (ER) in this context is the identification of reading errors made by another individual. A total of 57 3rd grade students, 22 males, and 35 females, ranging in age from 7:11 to 9:1 participated in the study. Participants were predominantly of Hispanic/ Latino descent with 6 African-American and 2 Asian- American student participants. Of the 57 participating students 23 were English Learners with California English Language Development Test levels ranging from 1 – 5. At risk criteria for the DIBELS measures were based on appropriate benchmark goals while at risk criteria for the ER task was defined as a score of less than 13 on the ER measure. Results indicated that the DORF measure had a slightly larger area under the curve (AUC) value followed by the ER measure and the Daze measure respectively. However, AUC value comparisons indicated there were no statistically significant differences between the AUC values obtained by any of the three measures. Limitations and future directions are discussed.
Keywords: Error recognition, early identification, reading, English Learners