After years on the educational “blacklist” it appears that ability grouping has returned with force, possibly as a direct consequence of NCLB. Ability grouping is the practice of having students work with students who have similar “ability,” as determined by their perceived achievement level or talent for a particular subject. Think the old school stereotype of “buzzards” vs. “eagles” reading groups (guess which group was the high-achievers?). Over 60% of fourth-grade teachers apparently group by ability for both reading and math, according to a recent survey.
I am not a fan of ability grouping. My rationale is that despite being a “common-sense” method for differentiating instruction to help student achievement, there is fairly good evidence ability grouping doesn’t work. Tracking, or assigning students to whole classes composed of students of similar ability, has been shown to be especially ineffective in increasing student achievement, for students of all ability levels (see pg. 89 of John Hattie’s book, Visible Learning for a discussion of this). The exception to this may be when high-ability students receive an entirely different curriculum specifically designed for them, which rarely is the case.
However, even grouping students within a single mixed-ability class by ability seems to be ineffective, compared with other possible interventions. Hattie calculates the effect size of grouping by ability within a class as d = 0.15, based on a synthesis of meta-analyses (studies that group together the results of many studies on the topic). To put this in perspective, that is roughly equivalent to the amount of progress a student would make over the course of a year, without any instruction whatsoever- some, but not much. There are numerous classroom interventions that have effect sizes three to four times this size (for example, spacing out practice over time, rather than massing it all at once). So, if you have to choose where to put your effort, ability grouping is not the thing to do. When you include the possible equity issues and nearly inevitable effects related to students’ perceptions of the situation (“smart” groups versus “dumb” groups anyone?), to me it’s clearly a bad idea.