Additive Effects of Contingent Rewards and Performance Feedback on Reading Performance Under Intensive Reading Instruction

Two different interventions were implemented with 22 students receiving intensive reading instruction. A repeated-measures ANOVA, graphic analysis, effect sizes, and raw score gains were used to examine the effects of 1) performance feedback only and 2) performance feedback plus contingent rewards on several reading variables: fluency, comprehension, self-reported interest in reading, and voluntary engagement in reading. Four 3rd-grade classrooms were assigned to treatment conditions. Students in both conditions received 2-2.5 hours of reading instruction per day.

All students completed assessments of fluency and comprehension twice per week during the treatment phase of the study. Students in the performance feedback only condition completed these assessments and received feedback about their performance. Students in the performance feedback plus contingent reward condition completed the same procedures but also received rewards contingent upon improvement over previous performance; students received one sticker for increasing their reading fluency score and/or one sticker for increasing their reading comprehension score. Stickers could be used to purchase backup rewards. Prior to, at the conclusion of, and four months following the conclusion of the intervention, all students completed measures of reading skill, reading interest, and a choice condition to assess voluntary engagement in reading.

A statistically significant main effect on oral reading fluency was obtained but no significant main effects were found for retell fluency, voluntary engagement, or self-reported interest. Furthermore, no between-subjects main effects or interaction effects were found between conditions and phases. Results based on graphic analysis of data, effect sizes, and raw score gains indicated that students in both conditions showed improvements in reading fluency and comprehension, with students in the feedback only condition making greater overall gains on the former and students in the feedback plus reward condition making slightly greater improvements on the latter. Additionally, graphic representations of data show differences between the conditions on measures of voluntary engagement and self-reported interest. On the voluntary engagement measure, students in the feedback plus reward condition made greater gains than the feedback only group. On the self-reported interest survey, the feedback plus reward condition decreased throughout the study whereas the feedback only group increased.