This report describes the findings from the evaluation study of Read for Real conducted during the 2004–2005 school year. In the summer of 2004, Read for Real school sites serving grades 4 and 6, distributed across urban, suburban, and rural settings, along with similar school sites that were not using Read for Real, were engaged to participate in the evaluation study. The latter sites are referred to as OTHER. A total of 13 schools, initially representing 457 experimental and 470 control students, agreed to participate in the study.
The Gates-MacGinitie Reading Tests (4th edition) were administered to students as a pretest in the fall of 2004 and as a posttest in the spring of 2005. An analysis of adjusted posttest scores (using pretest scores as a covariate) indicated that Read for Real students had higher mean scores than OTHER students for comprehension, vocabulary, and total score. The contrasts for comprehension and total score were found to be significant at the .05 level; however, the contrast for vocabulary was not.
Entering the independent variable of grade level into the basic mathematical model did not change this finding, indicating that Read for Real did not have a differential effect at grades four and six. However, when locale (i.e., urban, suburban, and rural) was entered into the basic model, a statistically significant differential pattern of effects for Read for Real versus OTHER students was observed. Caution regarding the interpretation for this differential pattern is suggested due to the unequal cell frequencies in the model and the particularly small number of subjects in urban settings.
The basic findings involving Read for Real and OTHER students were interpreted using the Binomial Effect Size Display. This indicates that Read for Real students would have a 7% higher passing rate than would OTHER students if assessments of reading comprehension were administered that have an expected passing rate of 50% in the combined population of Read for Real and OTHER students. Similarly, Read for Real students would have a 7% higher passing rate than would OTHER students if assessments resulting in a composite score of comprehension and vocabulary were administered that have an expected passing rate of 50% in the combined population of Read for Real and OTHER students.
© Zaner-Bloser, Inc. All Rights Reserved.