Re: [Phys-l] differentiated instruction

["John Clement" <clement@hal-pc.org>]

> The idea behind my second question is this.  Say we accept that there are
> many ways to teach.  And that each student has a method in which they are
> most able to learn.  It might be the case that more physics will be
> learned if we adjust the methods to match the student's preferred mode.
> But it might be BETTER to continue to present many methods to all (call it
> "shotgun" if you  like) and then help each student to get something from
> each of them.

There are certainly many ways to teach, but I submit that it is a fallacy to
say that there are many ways to learn.  It is true that there are many ways
to "memorize" and people do use all kinds of individual tricks to do this.
But to learn something like physics the research shows that certain methods
of teaching in general produce significantly lower results.  Students have
to integrate the ideas, and make connections.  They have to think things
through.  They have to use proportional reasoning.

Let us take an example:
It is possible to explain an algorithm and then have students practice it.
Or it is possible to give students problems, let them develop methods in
groups, and share methods.  Finally an efficient algorithm can be explained.
The latter method produces transfer, while the former does not, except
possibly in an occasional student.  This has been shown by Schwartz, and
also has been shown for low performing elementary school students.

It is now known that getting students to solve kinematics problems
graphically without algebra first produces more reliable problem solving,
and greater understanding overall.  In addition to gain on the FCI there is
strong gain on the TUGK.

The other problem is the "preferred" mode.  Students prefer to be told what
to do, and just repeat that on tests.  This is the kiss of death for
understanding or good problem solving.  Students do not really understand
how they learn.  This came home to me when I realized that working with
others improved my grades, but when I worked alone my grades were lower.

Essentially it looks like differentiated learning is differentiated spoon
feeding.  Not only is it impossible in HS, but it is destructive of real
learning.  The problem is that often one must show that the instruction is
differentiated to maintain a job.  My suggestion that using graphs will look
different to an administrator from doing algebraic solutions, so you can
claim it is differentiated even though you are requiring students to
eventually deal with all representations.  It is not scatter gun, but rather
very targeted.  But to an administrator it might look OK.

Much of this comes from the observation that there are "multiple
intelligences".  This is just noting that different people become expert in
different ways.  There are musical geniuses who may not know any physics.
There are literary geniuses who sell many books, or whose plays survive for
centuries, but who harbor all of the common misconceptions about science.  I
always cringe a bit when I see the common artistic depictions of the new
moon with the crescent standing straight up.  And so on.  But does this mean
one must use music to teach physics, literature to teach physics...  No.
Musical jingles are for memorization, and do not get across the concepts.
Students have to all be able to draw the diagrams, parse the graphs, explain
what is going on and lastly use some equations.  Ideally they should use
both.

I contacted the author of this idea, and he essentially disavowed that the
idea of multiple intelligences could be used to shape instruction.  As I
recall he said that students have to be given challenging tasks.

Basically much of this comes from pop psychology, and not from real
cognitive science research which looks at how students learn math and
science.  So districts put teachers through "color training" to recognize
different modalities of learning...  Private schools mandate colored paper
for students even though the developmental pediatricians will tell you this
is unproven.  There is way too much of this type of thing aiming at a quick
fix!

Can you give differentiated tests?  Does the state give differentiated
tests?  Are there differentiated versions of the SAT?

John M. Clement
Houston, TX