The idea (I'll resist the strong temptation to put the word idea in scare quotes for the time being) is that teachers will tailor (or differentiate if you will) their instruction to the individual needs of each student. And by doing so, the theory goes, the students will learn more.
But is there any empirical support for such a theory. Let's set the wayback machine to 1986 and listen to the conclusion set forth by James Brophy in a paper he presented that year:
Research has turned up very little evidence suggesting the need for qualitatively different forms of instruction for students who differ in aptitude, achievement levels, socioeconomic status, ethnicity, or learning style.Nothing has changed in the intervening two decades to change that conclusion. In fact, by 1986 we already knew of one instructional program which was effective and which taught each student the exact same material. Engelmann, the creator of that program, clarifies Brophy's conclusion and how it relates to instructional design:
Some kids will need more practice exercises than others. Some will take more time to accumulate the skills needed to enter a particular program. But if the program does an effective job of communicating with the kids, showing exactly what to do and providing adequate practice for the kid who needs more practice, the program is a good program for all kids who have the skill assumed at the beginning of the program.In 2006, most educators would disagree vehemently that learning styles and individual differences have a relatively minor effect on kids' academic performance. They've turned their classrooms into a chaotic mess by trying to provide different instruction to each kid.
In engineering, we call this a kludge--an inelegant solution to a problem. To make matters worse, the problem is self imposed. The problem is heterogeneous grouping of students. Throw all the kids into the same classroom and try to teach them the same stuff. Apparently, it doesn't work too well. Duh.
So, instead of scrapping the whole idea of heterogeneous grouping and going back to (flexible) homogeneous grouping, educators, instead of admitting wrong doing, are attempting to fix the problem by applying a differentiated instruction band-aid. This let's them keep on doing what they've always been doing by making superficial changes.
I'm surprised they didn't call it balanced grouping.
Update: I found this post at JIS Topics that discusses their school's recent adoption of Differentiated Instruction.