Re: [Phys-l] unbiased experiments +- index of refraction

[John Denker <jsd@av8n.com>]

On 05/12/2009 10:14 AM, Philip Keller wrote:
> Are we sure that there is NO educational value in "verifying" labs?

Nobody is claiming that.

> That's a strong claim, and if it is true, I am troubled to hear it
> because I sure do a lot of them in my first-year high school physics
> classes.

That's the point:  the pendulum is so very far unbalanced in
one direction that we can move a long ways in the other direction
before we achieve balance, let alone start worrying about unbalance
of the opposite kind.

> For example:
>
> We use motion sensors to get velocity data to see if momentum is
> conserved in a series of collisions.  We also check if energy is
> conserved, and if not, we calculate the fraction that is "lost" in
> the collision.  Is this a pointless exercise for a first year class?

That depends on how it's done.

There is such a thing as a _null experiment_.  Some of the greatest
experiments of all time have been null experiments.  The students
ought to learn how to do such experiments, and why they are important.
They weren't born knowing this, so you have to spend some time and
effort laying the groundwork.

Suppose the result of the measurement is A ± B.  In the conservation
experiment, as is often the case in null experiments, B is more 
interesting than A.  This B tells you how accurately you can measure
momentum, using the given apparatus and techniques.  Students who
find the conservation result (A) unsurprising are likely to find the
uncertainty (B) surprisingly small.

> As I have argued on other occasions, teaching involves far too many
> pedagogical judgments to investigate each one with a FCI-linked
> research project.  

That's for sure.