UC Riverside

Growth rates generated from a single probe per measurement occasion versus three probes and taking the median were compared by examining the scores psychometric characteristics. Students who were struggling in reading were monitored with DIBELS Next progress monitoring passages during the academic year. Data were collected from six elementary schools by the primary researcher yielding seven weeks of data for 219 students. A paired-samples t-test indicated a statistical difference between the single probe and median probe model growth rates (p < .01). However, predictive validity analysis indicated both were significant predictors of future reading performance (p < .05). There were also minimal differences between the standard error of the estimates and R-squared values produced. Based on the results generated, it is suggested that a single probe progress monitoring approach would not compromise the predictive validity and utility for making important educational decisions (i.e. guide instructional decision-making).