I had lunch last week with one of my high school teachers. One of the topics that came up over guacamole was whether heterogeneous or homogeneous grouping is better, i.e. should classes be mixed ability or should “fast” and “slow” students be tracked separately. After a quick laugh–we’re both convinced that heterogeneous grouping is the way to go–we moved on to the tacos and burritos.
I should explain that in my school (this was a few years back), every class was mixed by academic ability, by income, and by race. Every classmate I know remains convinced that this is the only way to go. But maybe this worked so well because we had such incredibly good teachers.
So here’s a bit of quantitative evidence that disagrees with my firm convictions about the benefits of mixing up students. Courtney Collins and Li Gan ask Does Sorting Students Improve Scores? in a new working paper that uses data from the Dallas public schools to compare test scores in schools which mix students up to scores from schools that emphasize ability tracking. The first argument is that it’s easier to teach to students of similar abilities. According to this argument all students benefit from being tracked with like students. The second argument is that having high achieving peers helps students. This argument suggests a second reason that high achieving students benefit from tracking–they’re entirely surrounded by other high achieving students. The flip side is that low achieving students lose out–they miss the leavening from more academically accomplished peers.
Collins’ and Gan’s findings are that both high and low achieving students benefit on net. This is true for both math and reading scores. Oddly (i.e., contrary to what their theory suggests), low achieving students get an even bigger improvement from tracking than do high achieving students.
I’m left wondering, what happens to kids who spend 12 years in schools with other kids who are just like they are? What happens when they graduate into the real, very heterogeneous world?