Although it might be true that... trying to identify individual learning styles, then trying to provide tailor-made styles of instruction to several different types of students, within one class simultaneously... is not practical, I'm afraid that most people indeed mean exactly this by "differentiated instruction."
If you have a variety of learning types in your class, and you simply try to include all students by varying your teaching style to cover all or most of the learning types, thereby hoping all students are served, this would be what I call the "shot gun approach." The shotgun approach does indeed work for some things, such as duck hunting, and that's why shotguns exist.
Your efficiency as a teacher goes way down if you are providing a particular style of teaching that no one needs. I don't see the shotgun approach as efficient. There is indeed the expectation that you will assess the learning styles of every student in your class, and attempt to tailor approaches to each and every one of them.
The Wiki entry on differentiated instruction says, "The first and most important step in differentiated instruction is determining what students already know so as not to cover material students have mastered, or use methods that would be ineffective for students."
Unless you have segregated classes, your students will not be homogeneous. Indeed, many elementary-school programs today specifically do not segregate students (although they did when I was in elementary school in the 1950s). If we don't have or can't have segregated classes, how can we be effective with a differentiated approach?
My wife teaches 5th-grade math to all 5th graders in Bluffton (about 80 students). She has spent a lot of time assessing students herself, and also talking to the 4th grade teachers to get their assessments of individual students. She is then expected to provide simultaneous instruction to a class of students with mixed needs. Especially poor or problematic students of course have IEPs (individualized education plans). Although the government only requires IEPs for students with disabilities, our elementary teachers somewhat extend that process for all students. They don't spend as much time on the non-problem kids, and IEP sort of becomes a GEP (group education plan) for the more normal students. but my wife does have two or three learning styles that she tries to groups and reach in addition to the IEP kids.
One big problem I see with this is my wife is burning out. This is a heck of a lot of work. She spends about 2 to 3 hours a day on lesson planning because she is planning for a couple groups, plus the IEP kids. Right now I make one plan for each class period. She makes multiple plans for each segment of the day. Goodness! Who long can she keep that up?
The bright side is that for the past several years her students' performance on the State 5th-grade math tests have been near the top. Last year Bluffton 5th grade was fifth-highest in the state. The cost is very high... not in money... but in the life of the teacher.
At the college level my classes used to be fairly homogeneous. Students who didn't have physics aptitude didn't take my courses, and maybe didn't even come to college. Now they are in my class because they want to be doctors or some lofty goal, but half of them can't remember whether sin is opposite over hypotenuse or adjacent over hypotenuse, and they don't understand calculus at all (even though they passed the course), and they didn't take physics in high school. But the other half is pretty good. But we don't have enough students for two sections, so they're all mixed in the same class. If I try to make this work as hard as my wife is trying to make her class work, I won't have a life.
Michael D. Edmiston, Ph.D.
Professor of Chemistry and Physics
Bluffton University
Bluffton, OH 45817
(419)-358-3270
edmiston@bluffton.edu