ABSTRACT
We compared children's gains in oral reading fluency after applying a standard fluency-building intervention to three training passages that differed in word overlap (high, low, and multiple exemplar) with an untrained generalization passage. Participants were 132 White and Hispanic third-grade children from two schools in the northeast and mountain west. Children were randomly assigned within classrooms to the three word overlap conditions, pre-tested on their assigned training and a common generalization passage, received a fluency-building intervention on their assigned training passage, and then post-tested on the same two passages. Regression analyses were conducted to examine the effects of word overlap condition on the children's fluency gains after controlling for pre-test fluency and classroom. Results revealed significantly larger priming and generalization effects for the multiple exemplar versus both the low- and high-word overlap conditions. Survival curves showed that a significantly larger proportion of children in the multiple exemplar condition survived as generalized responders at all generalization levels relative to the other two conditions. Implications for assessing and promoting generalized oral reading fluency in response-to-intervention models and directions for future research are discussed.