Accelerated Reader in a primary school: An evaluation of time spent on classroom implementation and student achievement.

摘要

The year-long study examined comparison groups of second grade students using the Accelerated Reader (AR) program 60 minutes versus 30 minutes per day. Three sources were used for quantitative data: the Dynamic Indicators of Basic Early Literacy Skills (DIBELS) test, the Standardized Test for the Assessment of Reading (STAR) Early Literacy test, and the Accelerated Reader (AR) software database. The results did not show a significant difference in the posttest scores of the students using the AR program 60 minutes, instead of 30 minutes, per day. These results raised questions about variations in individual teachers' implementation of AR activity time, as well as the differences in the tests used in this study and those used in many other research studies on the AR program. As expected, the 30-minute and 60-minute groups of students in the lowest quartile made greater improvement than those same groups in the highest quartile. This gain was not significant in the oral reading section of the DIBELS. However, it was significant in retell fluency and word use fluency areas of the DIBELS. In addition, the gain was significant for all sections of the STAR Early Literacy Test. The anticipated higher gain for the lowest quartile could be attributed to the substantial amount of potential improvement for students in this group and extensive school personnel assistance to help them acquire basic necessary skills in reading. Also, the highest quartile had a smaller amount of room for improvement from pretest scores. The study showed a moderate association (r = 0.60) with DIBELS oral reading and the number of AR books read by the students. There was no correlation between the number of books read and the retell fluency or word use fluency sections of the DIBELS. Additionally, there was no correlation between the number of books read and all areas of the STAR Early Literacy Test. In analysis for a correlation between the average percent correct on AR tests with the DIBELS and STAR Early Literacy Test, there was no relationship shown in DIBELS retell fluency, word use, or the STAR scaled score. The study found a moderate correlation (r = 0.52) between the average percent correct on AR tests and DIBELS oral fluency, while all other areas of the STAR Early Literacy Test had a moderate relationship (r = 0.60) with the AR test average. The surprise in this study was that students in the 60-minute group did not exhibit significantly higher test scores than those students in the 30-minute group.