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]

[Phys-L] Re: Human Error?

I'd like to second some of this. In introductory physics, we're usua=
lly asking the students to replicate some well-known result such as t=
he acceleration of a ball down an inclined plane. There is nothing ne=
w or exciting to be learned here, and the students know it. Therefor=
e the best they can do is to try to get the accepted value. Usually=
, their experimental technique is too poor for this to occur. We ca=
n, however, introduce the concept of an "error bar" and what it means=
to "agree" experimentally. Besides being totally unfamiliar with th=
e concept of quantitative error analysis, most students are amazed to=
hear that the accepted value has error, too. It takes some effort=
to steer students away from the vague and automatic "human error" to=
ward a more thoughtful, insightful, and quantitative approach.

One way to do this is to find some variation on the standard experime=
nt for which there is no accepted value - finding the spring constant=
of stretchy toy snakes, for example. Students must be taught what t=
o do when there is no "accepted" value. They just don't know how to =
cope with that. The have never experienced the kind of experimenting=
that's done to answer a question. (Which, of course, is the whole p=
oint of doing experiments!)=20

Part of the difficulty is that many students have very little concept=
of "mechanism." If you ask 'em how it works, they say, "push that b=
utton." In their common experience, either something works perfectl=
y or it doesn't. If it can't be restored to functionality by recharg=
ing or replacing the battery, it gets thrown out and replaced. Nothi=
ng can be repaired. Everything is purchased - nothing is made at hom=
e. Even if you know what's wrong, you can't do much about it. So mo=
st people never bother to find out what's wrong with a malfunctioning=
device - they just replace it. Students literally have no clue wher=
e stuff comes from or what makes it work. Therefore the concept of di=
gging deeper into an experiment in order to figure out why it's not w=
orking perfectly is also foriegn. Even my calculus-based students d=
idn't know that the "green thing" (the circuit board) contains resist=
ors and capacitors and is responsible for the functioning of their co=
mputers and cell phones. So we're fighting an uphill battle, everyon=
e.

Vickie Frohne

-----Original Message-----
=46rom: Forum for Physics Educators on behalf of Jack Uretsky
Sent: Sat 4/30/2005 12:37 PM
To: PHYS-L@LISTS.NAU.EDU
Subject: Re: Human Error?
=20
Hi all-
I'll add to John's remarks that there has been too little
discussion of why we bother calculating (or estimating) uncertainties=
.
The answer is that it is often important to know when two experiments
agree (or disagree) with each other. This decision may lie at the he=
art
of identifying a new physical phenomenon, or deciding which experimen=
tal
result to use in predicting the outcome of a new experiment. There h=
as,
for example, beeen a lively (and occasionally acrimonious) debate ove=
r
whether the latest measurement of the muon g-2 disagrees with theory
-thereby indicating the influence of "new physics" - or agrees within=
the
experimental and theoretical uncertainties. Part of the theoretical
uncertainty is the value of the fine structure constant, which is obt=
ained
=66rom a separate experiment.
It may be useful to compare with "accepted values" if those v=
alues
are accompanied by their own uncertainties, and the student is requir=
ed to
discuss the extent to which his/her result is in agreement with the
accepted value.
Regards,
Jack





On Sat, 30 Apr 2005, John Clement wrote:

This is where some good teaching using reformed pedagogy can be of =
value.
For example in Modeling the whole point of view is that the student=
s are
creating a model for what they observe, and that this model can be =
modified
when necessary. The whole orientation is around what they observe =
rather
than what a book says. This point of view runs through virtually a=
ll of the
PER curricula, and should be a good antidote to the idea of accepte=
d values
being used to calculate errors.

Unfortunately the idea of calculating experimental-accepted value i=
s
actually promoted by some state curricula and is in many textbooks,=
and in
teacher's notes. If you want to fight this at its source then you =
have to
get involved in the state curriculum committees.

John M. Clement
Houston, TX



As an aside, I do have this same problem with juniors and seniors=
in my
experimental
physics course for physics majors at our university. They think =
that
their only
discussion in the lab report should be about "error" and often ig=
nore the
physics of the
experiment. They do have this belief that there is always some "=
accepted
value" and
that there is no way they can make a measurement without "human e=
rror."
In one
recent report, a student claimed, "The error in this case cannot =
be human
error because
all of the data was measured using a computer."

The experiment in question is the Kater pendulum, where our objec=
tive is
to measure
the local value of the acceleration due to gravity to the fourth =
decimal
place. We get
this accuaracy by measuring over 125 periods at each of 32 bob po=
sitions;
and by
measuring the distance between the two fulcrums to within 100 mic=
rons
using a
precision cathetometer. Even though it is communicated to the st=
udents
that our goal is
a precise measurement of g, some of them still want to compare it=
with the
book value
of 9.81 m/s^2 and calculate a percent error, even though it is cl=
early
stated not to
compare the measured result with any other number. They are aske=
d to do
an
uncertainty propagation to prove the precision of the measurement=
. Almost
none of the
students attempt this, even though they were given practice probl=
ems at
the beginning
of the semester on how to do it.



--
"Trust me. I have a lot of experience at this."
General Custer's unremembered message to his men,
just before leading them into the Little Big Horn Val=
ley
_______________________________________________
Phys-L mailing list
Phys-L@electron.physics.buffalo.edu
https://www.physics.buffalo.edu/mailman/listinfo/phys-l