Re: calculus-based physics final exam

["John S. Denker" <jsd@MONMOUTH.COM>]

At 21:50 -0400 4/16/03, Aaron Titus wrote:
>>
>> I'm not sure being too hard is a problem. I've heard that a mean near
>> 50% is actually desirable (in testing and measurement "theory"). The
>> main thing that I want is the ability to compare my students' scores to
>> those of students at other institutions.

On 04/16/2003 11:36 PM, Hugh Haskell wrote:
> I don't have any problem with a test whose mean is about 50%, but the
> students scream like wounded pigs over them. They are all used to
> tests where the average is at 80% or higher, so when they get one
> with a mean of 50%, they go bananas.

This raises deep issues.

Everything that happens in class, tests included,
advances multiple agendas, some more than others.

The most obvious agenda is to provide plain old
information about the subject matter.

Another important agenda is to provide the students
with motivation.

Another important agenda is to provide them with
"pacing".  It's not possible to learn everything
at once.

Another important agenda is to instill in them
some confidence that they can handle the material.

Another agenda is to evaluate them, so that they
know where they stand, and so you and perhaps
others can know where they stand.

==

It is only a slight exaggeration to say that any
students who have sufficient endogenous motivation,
pacing, and confidence don't need to come to your
class at all ... they can learn the material straight
out of the book.

If you think that the predominant purpose of the
test is to evaluate the students, then sure, go
ahead, set the median score at 50%.....

But realize that you are paying a high price for
this evaluation.  You are probably losing most of
an opportunity to convey information about the
subject matter.  What's worse, you are conveying
misleading information about the pacing, providing
negative motivation, and reducing their confidence.

One of the first things I ever did as a teacher
was to assign a hard test.  I was thinking about
my needs for evaluation/information.  I wasn't
thinking about the students' needs.

Eventually I discovered there's a lot more to
teaching than "test and measurement theory".