- Many thanks, John, for that insight and for doing the work of writing it
and sharing it.
- I will share it with some of my relatives.
- And thanks for your work in improving the website resistance to
miscreants.
Bill Norwood
U of MD at College Park
On Feb 15, 2017 3:23 PM, "John Denker" <jsd@av8n.com> wrote:
Hi --
In an attempt to raise the signal-to-noise level in this channel,
let me tell a story.
Once upon a time a fourth grader conducted a science-fair
experiment that went like this:
Purpose: Find out what ants like to eat. [a]
Hypothesis: They like sugar more than anything else. [b]
Method: Set out four plates, one with meat, one with bread,
one with soft brown sugar crystals, and one plain empty
plate as a control. [c]
Repeat that at four different anthills. Record the number
of ants on each plate, versus time. [d]
Observation: At one site, the ants swarmed all over the meat.
At another site, the swarmed all over the sugar. At the
other two sites, there were very few ants on any of the
plates, and the empty plate had as many ants as any of the
others. See the data tables and graphs. [e]
By the way, they looked like four different kinds of ants.
They were different sizes and shapes and colors. [f]
Conclusion: My hypothesis was wrong. Next time I will do
the experiment with all the same kind of ants, so I can
get better data. [g]
=============================
That's not quite verbatim, but close. You can't make this
stuff up.
Take a moment to imagine my reaction. Just imagine.
=============================
I buttonholed each of the teachers and reviewed the experiment
with them. I said that
[a] was excellent,
[c] was excellent,
[d] was excellent,
[e] was excellent,
[f] was brilliant,
whereas
[b] and [g] were tragic. They epitomized some deep-seated
problems with the way science fairs are typically conducted.
The kid did everything right, right up to the very last step,
but was allowed -- essentially required -- to draw exactly the
wrong conclusions.
I managed to track down the kid. I explained: You carried out
the experiment exactly right. This is what science is supposed
to look like. This is why we do experiments: to find out stuff
we didn't already know. Very very often we find something what
wasn't exactly what we were looking for. This happens so often
that there is even a specific word for it: Serendipity. Look
it up.
This is what happens when you're smart and keep your eyes open
and put yourself in situations where something interesting might
happen.
As for the poster: Please rewrite the conclusion. You should
be celebrating this result, no apologizing for it. Rewrite it
to say you made a serendipitous discovery. You learned that
different kinds of ant like different foods.
Secondly, get rid of the hypothesis. Eliminate it not just
from the conclusions but from the whole poster. Eliminate the
word and the concept. It's the wrong word attached to an unwise
concept. A hypothesis is not a guess, and you shouldn't be
guessing anyway. One of the great perks of being a scientist
is that you don't have to guess. At step [b] you should be
making a list of /things that could happen/. Things plural.
This is required for safety and required by common sense. It's
a lot more scientific than guessing. It's obvious from your
methods that you were considering at least four different
/things that could happen/ -- considering them all simultaneously,
as you should.
So the point remains, operationally you did everything right.
You just need to tune up the poster to describe it in more
scientific terms, so that you draw the right conclusions.
As for the last part of the conclusion: Next time do *not*
restrict it to one kind of ant. Do it with *more* kinds of
ant! Science is not a guessing game. The idea is to learn
something. If you want to check for reproducibility, it is
OK to repeat the experiment at two or three different anthills
per species, but the signal you got is so strong that this
is hardly worth the trouble. In any case, don't give up on
cross-checking one species against another. Take close-up
pictures, so you can more easily tell which ants are which.
=============================
Last but not least: Judging a science fair is not a very
productive activity. The kids don't need judges; they need
mentors. Almost every science-fair experiment has significant
flaws that could have been fixed in a femtosecond if they had
been noticed early enough.
_______________________________________________
Forum for Physics Educators
Phys-l@mail.phys-l.org http://www.phys-l.org/mailman/listinfo/phys-l