The Labeling Effect of Ability Grouping

Abstract

To determine whether assigning ability-level labels to randomly grouped mathematics class sections has any effect upon their academic achievement, a group of one hundred enlisted airmen at the United States Air Force Academy Preparatory School were assigned randomly to five simulated ability levels. Section grade averages were compared using the two-tailed t-test. For all eight data sets derived from the experiment, differences between means of top-label and bottom-label sections were significant at the 0.1+ percent level. For twenty-nine comparisons of means of adjacent ability-level label sections, upper-label sections exceed lower-label sections in all but three cases. For these three cases the differences were negligible. This shows that a labeling effect undeniably exists in ability grouping situations.