| Chronology | Current Month | Current Thread | Current Date |
| [Year List] [Month List (current year)] | [Date Index] [Thread Index] | [Thread Prev] [Thread Next] | [Date Prev] [Date Next] |
On 09/09/2011 05:30 AM, Joseph Bellina wrote:
First if one is to design an experiment then you have to have some
vision of what it is you want to find out.
That's true as stated ... but it doesn't detract from the point I was
making. When I toss a coin, I have some vision that it will come up
heads *or* tails. I don't need to predict the specific outcome. The
rule is to consider the entire range of plausible outcomes.
-- Some vision: yes.
-- Ultra-specific prediction: no.
This is a big deal in connection with "science fair" projects. Go to
the next fair and look around. Count the number of posters that say
"my prediction was true" or "my prediction turned out to be false"
... when the experiment was basically a coin toss that could have gone
either way. Whenever I see that I have to bite my lip. I want to
scream "this is not science". It's a travesty of science. It makes
a mockery of common sense. The Boy Scouts are more scientific than
the science fair, in the sense that when planning for a campout they
hope it won't rain but they plan for the possibility that it will.
The saying is, hope for the best but plan for the worst.
In manufacturing you often have a situation where you want the finished
products to be extremely uniform and "predictable". On the other hand,
even in that situation, if you go one or more steps upstream of the
finished product you find all sorts of quality-control inspections and
feedback steps that explicitly consider the *entire* range of possibilities,
including out-of-spec as well as within-spec items.
This applies a thousand times more to engineering research (as opposed
to mass production) and a million times more to scientific research,
where you are absolutely not required to know exactly what's going to
be the result. Yes, you know in general terms, just not in specific
terms.
Since baseball analogies seem to be in season, consider the legend of
Babe Ruth calling the shot. Maybe he did, maybe he didn't, but the
point remains that he wasn't required to do so. Hitting a home run is
hard enough without having to specify which seat it is going to land
in.
-- Some vision, yes: I'm going to swing as hard as I can, and with a
little luck it might be a home run.
-- Ultra-specific prediction, no: Calling the shot would be pointless
and irrelevant. If it's a fair ball and it lands in the seats, it's
a home run, no matter which particular seat.
Also aviation analogies seem to be in season. Before taking off, you
want to predict with confidence that you will be able to land safely.
This includes not running out of fuel enroute. I can calculate rather
precisely how much fuel "should" be required, but I make sure there's
a lot more than that on board. There are multiple rules about this.
One of the federal regulations requires enough fuel to fly to the
intended destination, then fly to the alternate, and then fly for 45
minutes more than that /at cruising speed/ (which corresponds to a lot
more than 45 minutes at holding speed). There is also a rule that says
if you are running low on fuel, due to worse-than-predicted headwinds
or whatever, you should land well before the intended destination ...
nevermind the alternate or anything beyond that. The latter rule is not
specifically codified in the federal regulations, but it's a darn good
rule anyway. The point remains: The broad outlines are predictable,
but the details are not.
-- Broad prediction, yes: Within five hours after takeoff, this airplane
*will* have landed somewhere.
-- Ultra-specific prediction, no: It won't necessarily be at the intended
destination. You *must* consider the entire range of plausible outcomes.
_______________________________________________
Forum for Physics Educators
Phys-l@carnot.physics.buffalo.edu
https://carnot.physics.buffalo.edu/mailman/listinfo/phys-l