My main objection is to the word "error" a number of responders have
pointed out that "uncertainty" would be a better word. And, I agree.
In may labs, I try to have students calculate "percent difference" between
two values and try to emphasize that this does not imply that their
answer is necessarily wrong just because there might be a high percent
difference. However, sadly, percent difference has connotations of error
even when I emphasize the contrary.
The student needs some way of comparing his answers to acceptable values,
but should not expect to get those values due to inabilities to control
all of the necessary variables adequately.
Students often think of physics as and "exact science," when it is really
just the "science of approximations." We usually can be as accurate as
we want, but do not have the wherewithall (especially financially) to be
as accurate as students sometimes come to expect.