"Seriously, when was the last time you saw a textbook say "Here are
two equally good explanations for the data"? When was the last time
a teacher asked students to defer judgment as to which theory is
better?"
One practical problem is that at the intro level it's not easy to find
situations where there are plausible models that nevertheless fail on all
counts, and where there are models that do quite well that are accessible
at the intro level. We found one such case study, that of formation of
sparks in air (chapter 21 in our textbook). Two spheres charged + and -,
bring them closer and at some point you may get a spark. The first
plausible model, one often suggested by students, is that the electrons
jump to the + sphere, but it's easy to see that the mean free path is
extremely short in air.
The next plausible model, again one often suggested by students, is that
the spark happens when the field gets big enough to yank electrons out of
the air molecules, thereby providing free electrons and positive ions which
can constitute a current, but a simple calculation of the approximate field
required to pull an outer electron out of an atom is about 1e11 V/m whereas
the observed field strength adequate to trigger a spark is about 3e6 V/m. A
student's first impulse is often to look for corrections, but fairly
quickly there's a consensus that it's going to be pretty tough to finagle
around 5 orders of magnitude.
With respect to both of these models, there's a glimpse of how science can
often rule out a proposed explanation even if no good explanation is
available. Or to put it more strongly, science is particularly good at
ruling out a model but can't be asked to "prove" that a particular model is
"correct".
A third model is one that students are unlikely to come up with themselves:
Suppose there is somewhere in the air a free electron (and ion), due to the
passage through the air of charged particles, either muons produced in the
upper atmosphere in nuclear collisions (which reach the ground despite
their 2 microsecond lifetimes due to relativistic effects), or charged
particles due to radioactivity in the neighborhood. If there is a field big
enough to accelerate an electron in one mean free path to a kinetic energy
sufficient to knock an electron out of an atom, you then have two free
electrons, then 4, 8, 16, ..... (an "avalanche") and the air gets heavily
ionized. It's relatively simple physics to estimate mean free path and
ionization energy, which leads to a prediction of 3e7 V/m for the critical
field necessary to cause a spark. Given the simplicity of the model (for
example, electrons that happen to go more than one mean free path could
gain ionization energy in a smaller field), students are comfortable that
we've gotten within an order of magnitude of observation.
There's a nice footnote to this tale. One can now ask, "What if the density
of the atmosphere were twice as great?" The quick intuitive response is
that the critical field is now smaller "because there's more chances for
something to happen" or some such. One then looks at what each of the three
models would predict. The first two models predict no change in the
critical field. The third model however predicts that the critical field
will be twice as big, because the mean free path is half as big. Indeed,
this is what is observed (and high-density gas is sometimes used as a good
insulator). Here is an opportunity to say that a really good model is one
that explains more than was originally asked of it.