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]

Re: right or wrong?

Please stay away from language like "right or wrong", the scientific
world is fuzzier than that. The issue always is "are these two values
close enough that I am willing to say they are the same or that they are
different." We are in an inductive mode, and our experiences and
evaluations are always contingent.

I believe that too many weak high school students are drawn to science
because they see it as a place where they can know they are right or
wrong. When they get to college, it appears that way as they live in the
artificial world of textbooks, but when they get in the lab it gets fuzzy
and they are in trouble.

cheers

On Mon, 12 May 1997, LUDWIK
KOWALSKI wrote:

On May 12, 1997 George Spagna wrote:

I don't think the original question was about the efficacy of using
percentages vs. units when reporting uncertainties. The issue is whether
the uncertainty is measured relative to the experiment, or relative to some
"right" answer.

The "formula" quoted in the original post is clearly telling the students
that the purpose of scientific experiment and measurement is to get the
right answer, reinforcing the notion that the content of our discipline is
one of memorizable facts rather than a process for examining the world
around us.

How can I teach, and grade lab reports, if I do not accept the idea that
"the purpose of scientific experiment and measurement is to get the right
answer"? In one setup the right answer may be 9.8 +/- 0.2 while in another
it can be 9.8 +/- 2 (or +/- 0.0002, if you a prospector). The answer does
not have to be quantitative but it is always either right or wrong. That
is why teaching physics is different from teaching artistic subjects, for
example.
Ludwik Kowalski

P.S. A philosophical issue of objectivity does not have to be involved in
this consideration. That why the term "accepted value" is better than "true
value". But that is a topic for another thread, if you wish.