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: Error Analysis





On Wed, 11 Sep 1996, Karshner wrote:

I sometimes wonder if college freshman are mature enough to understand
error analysis, and feel very lucky if I can get them to record answers plus or
minus some error after my course. For some it seems obvious and others I think

I disagree. They may not be able to grasp how the standard deviation is
related to experimental uncertainties (I prefer uncertainty to error in
this context.), but they can understand that measurements are not exact
and develop a concept of accuracy/uncertainty of a measurement. Pass a
meter stick around the room and ask the students to find its length in
1/16s of an inch, using foot rulers, repeat measurements of the period of
a pendulum with stopwatches or perform some other simple experiments.
Then, after tabulating the results, ask "What is the best estimate?",
"How close do you think it is to the actual value?", "How close would you
expect to come if you repeated the experiment?" or similar questions.
After some thought and discussion, most students will be able to give
reasonable answers. They may even be able to grasp the idea that the
average becomes more reliable as the number of measurements increase,
although that is more questionable.
Al Clark.