Mastery learning and the decreasing variability hypothesis

Performance on successive units of achievement in graduate classrooms using mastery learning procedures was used to test two variations of Bloom's decreasing variability hypothesis—namely, that under the favorable conditions of mastery learning, differences in faster and slower learners will decrease over successive units, leading to (a) smaller variances on successive units and (b) smaller correlations between an initial measure of aptitude and achievement on successive units. The data from the four classrooms studied do not support the decreasing variability hypothesis; rather, they show no change over time.