Interdependent Group Contingencies with Randomly Selected Components Applied to Class-wide Performance in the Accelerated Reader Program

The Accelerated Reader (A/R) program is a computer-based reading program employed by numerous school systems that provides comprehension quizzes over 10,000 books that have been rated for difficulty level. The quiz information is used to provide feedback to educators. The A/R program is designed to incorporate the elements of choice and reinforcement. Despite the research that supports the use of choice and reinforcement to increase academic behaviors, the A/R reading program has not been successful in increasing the sustained silent reading behaviors of all students enrolled in the schools using this program. The use of group contingencies has been effective in increasing academic behaviors. The current study was designed to extend this research by investigating the effects of a group contingency on the reading behavior of 4th grade students. Thirty-two students across three 4th grade intact classes participated. The students continued to read A/R books and take quizzes as was the current practice. A multiple baseline design was employed to implement an interdependent group contingency with randomized components in each classroom. The intervention involved providing the entire class with a group reward based on the class-wide performance on the A/R quizzes. At the end of each week, a randomly selected criterion was compared to the class performance on the number of A/R quizzes passed. A randomly selected reward was implemented if the class performance exceeded the criterion. Visual analysis of time-series graphs provides some support for an increase in A/R quizzes taken and passed immediately after the intervention was applied. This increase was not maintained. Only one class showed a clear increase in percent of comprehension questions answered correctly after the group contingency was applied. Across all three classes there were no clear changes in reading level of books selected after the intervention was applied. When students were divided into ability groups based on their average baseline performance, the lowest performing students exhibited a statistically significant increase in quiz performance (i.e., quizzes taken, quizzes passed, percent correct, and book level). However, the middle and high performing students did not show any statistically significant change