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: what good is "percentage error"?



Hello--

I use the percent error calulation in my labs. It seemed (though I may
change my mind) harmless enough and an easy way to compare the
experimental result with some accepted value (which my students must look
up in the literature). However, I don't grade based on it, and I
only accept that particular calculation when there is some external
standard to compare to. If my students are comparing a calculated
quatity with a measured quantity, or two different calculated or measured
quantities, they calculate a percent difference.

What I do base my grade upon is a discussion of sources of error. I
expacet them to think through the experiment and determine (to the best
of their ability) what may cause error in the results. I also expect
them to think through what effect the source of error would have on the
results. For example, in a projectile motion lab, they predict a landing
distance based on a measured velocity off a horizontal table. Most will
correctly identify a source of error as that of the neglecting of air
resistance. However, often they fail to consider what that does. If
your ball went farther than predicted, then air resitance can't account
for that error entirely. I try to get them to see that there is more
than one source of error in the experiment. Sometimes the error combines
in such a way as to give a really bad result. Sometimes, the error
combines in such a way as to give you the desired result (or accepted
value). Again, with projectile motion, does the fact that the ball
lands exactly where predicted mean there is nor air resistance? Of
course not.

I think the important issue here is what your grading is teaching your
students. If you grade based on how small the "percent error"
(difference or whatever you call it) is, you teach your students that there
is a correct answer in science. If you ask them to delve further into
what there calculations mean, where the error may come from, and what
effect it may have, then you teach them that there is not necessarily a
"correct" answer, but that method is important and that there is an
"acceptable" answer.

Pardon me whle I duck the incoming responses :).

Mariam




+++++++++++++++++++++++++++++++++++++++++++
Mariam W. Dittmann
DeKalb College
Clarkston, GA

mdittman@dekalb.dc.peachnet.edu

+++++++++++++++++++++++++++++++++++++++++++