Mike Schmoker has this commentary in EdWeek:
Differentiated Instruction went on to become one of the most widely adopted instructional orthodoxies of our time. It claims that students learn best when (despite some semantically creative denial) grouped by ability, as well as by their personal interests and "learning styles."
I had seen this innovation in action. In every case, it seemed to complicate teachers' work, requiring them to procure and assemble multiple sets of materials.
I saw frustrated teachers trying to provide materials that matched each student's or group's presumed ability level, interest, preferred "modality" and learning style.
He couldn't find any empirical evidence that DI worked. Meanwhile, when he shared his observations, he got two reactions.
...When I shared these reasons with educators, many were glad to hear their suspicions affirmed. They had often been required to integrate DI into all their lessons-against their best instincts-as the program morphed, without any reliable evidence of its effectiveness, into established orthodoxy.
Others, however, were angered by any criticism of DI. Their reactions stopped some of my presentations dead in their tracks. These educators, and their districts, had invested enormous amounts of time, treasure, and hope in this pedagogical approach.
[MG note: Don't get confused. There's another eduterm, Direct Instruction, commonly called D.I. In this context, the D in "D.I." = Differentiated, not Direct.
So what does work? Schmoker writes:
First, we need coherent, content-rich guaranteed curriculum...
Second-and just as important-we need to ensure that students read, write, and discuss, in the analytic and argumentative modes, for hundreds of hours per school year, across the curriculum. We aren't even close to that now. All students should be reading deeply, discussing, arguing, and writing about what they read every day in multiple courses. We can do this: Consider that students spend about 1,000 hours per year in school.
Third...good lessons start with a clear, curriculum-based objective and assessment, followed by multiple cycles of instruction, guided practice, checks for understanding (the soul of a good lesson), and ongoing adjustments to instruction.
Thanks to the British educator Dylan Wiliam and others, we now know that the consistent delivery of lessons that include multiple checks for understanding may be the most powerful, cost-effective action we can take to ensure learning. Solid research demonstrates that students learn as much as four times as quickly from such lessons.
Nothing rivals these three considerations. Mountains of evidence proclaim their centrality. They should, therefore, be education's near-exclusive focus, our highest priority for at least a period of years-or until they are satisfactorily and routinely implemented.