Classroom Reinforcement and Learning: A Quantitative Synthesis

AbstractTo estimate the instructional effects of rewards or positive reinforcement, both primary and secondary, on classroom learning, 102 effect sizes were calculated from statistical data in 39 studies spanning a period of 20 years and containing a combined sample of 4,842 students in 202 classes. The mean of the study-median effect sizes is 1.17, which suggests average percentiles on learning outcomes of 88 and 50, respectively, for positively reinforced and control groups. Contrary to previous theory and opinions, the strong effects of instructional reinforcement appear constant across grades (kindergarten through college), socioeconomic levels, race, private and public schools, and community types; but tangible reinforcers may be slightly more effective than intangible ones; and girls and students in special schools and classes may be slightly more affected by reinforcements than other students.