Conference rusure::math

    I taught second semester calculus about a year ago and found that
    giving partial credit was great for the students and very time consuming
    for me.  Partial credit has been my norm in teaching but with calculus
    there can be problems being fair.  I decided my students would rather
    have the reinforcement for getting something right and not worry
    that much about whether I met the standards of Solomon.  I believe
    I was right and my extra work was justified by the results.

    Sometimes not giving partial credit is laziness and sometimes it
    is perfectionism.  In either case, I don't consider the environment
    the best for a subject containing many new concepts and requiring
    the student to build rapidly on the new concepts.
    
    In short, the class may be frustrating and not a happy learning
    experience.  Of course, all of the students will have the same handicap
    on tests etc. so the instructor should be evaluated on other strengths
    if any.  Not my cup of tea and I wouldn't like the class but it
    could be OK.
    
    Chuck

    I don't think it makes much difference.  Either way, the teacher has to
    adjust the difficulty of the test so that some students get a good
    portion correct and some students get a good portion wrong.  The
    purpose of the test is to measure ability and anything which scores the
    ability being taught while giving different scores to people with
    different ability is a suitable measurement.  (Then the teacher also
    has the job of standardizing the measurement.)
    
    I think it would be wrong to call requiring a student to get a problem
    correct an injustice.  The ultimate goal of teaching calculus is for
    the student to be able to do calculus.  If you're ever going to use
    calculus, say in engineering a bridge, you've got to get the right
    answer.  There isn't any partial credit if the bridge falls down
    because of an arithmetic error.  I don't see anything wrong with
    requiring a student to demonstrate that they can handle the entire
    subject before giving them a grade certifying that. 
    
    
    				-- edp 

    re .2
    
    I would be careful of these statements.  It is well known that calculus
    is used by engineering departments as a washout course.  I consider
    this silly but it is a fact of life.  (Calculus flunk out rate
    exceeds 50%.)  Not all teachers of mathematics feel the way I do
    about this state of affairs (most do, however) so one finds calculus
    teachers in sympathy with the engineering departments once in a
    while.
    
    Another point I should like to make is that not all of us in mathematics
    can do arithmetic - I was a consistent "C" student in arithmetic
    in grade school.  I did, however, fully understand that there is
    no correlation at all between arithmetical ability and mathematical
    ability.  On top of this, several excellent mathematicians I know
    cannot even make change!  They would never have completed their
    undergraduate math if arithmetical accuracy had been required of
    them.
    
    Insisting on exactness of numerical answers favors students good
    at that.  As far as getting the right answer in engineering is
    concerned, that is not a problem in any case.  I usually work such
    problems several times to debug the answers. 

    Chuck

    The best calculus instructor I ever had instructed the class to
    put any numeric answers in the form of fractions i.e. 3/17, or
    22/7, etc.  (He would NOT have considered 6/34 or 44/14 as a 'good'
    answer, however).
    
    I think I remember him saying that he found it hard to do arithmetic
    also, and that one could always resort to a calculator if necessary,
    but he was damned if HE was going to calculate decimal answers to
    his problems.
    
    I also seem to remember that MOST of his examination problems
    eventually resolved to extremely simple fractions.
    
    As a result, I do a lot of mental calculations in the same manner,
    now.

	In one memorable electrical engineering class, when we turned
	in problem sets, we were required to include a check.  Without
	a check, even if our answer and work were right, we didn't
	get credit!

	For example, suppose the problem were to solve

		x^2 - x + 12 = 106/9

	You could do the quadratic formula on your paper, and get
	2/3 as the answer.

	But no credit unless you showed a check too !  For example, you
	could show:

		(2/3)^2 = 4/9

		4/9 - 2/3 = -2/9

		-2/9 + 12 = -2/9 + 108/9 = 106/9

	I found this course VERY helpful, since in future work I had
	learned to do checks and hence greatly reduce the chance of
	having errors.

/Eric

>	In one memorable electrical engineering class, when we turned
>	in problem sets, we were required to include a check.  Without
>	a check, even if our answer and work were right, we didn't
>	get credit!

	Would a check for $10.00 suffice?

	When I was ten years old, my maths teacher gave occasional revision
tests in class.   There were a number of idiosyncrasies in the marking scheme
he used.

	(1) The number of points awarded for a question was equal to the number
of pupils who got the question *wrong*.

	(2) To a question such as "How many edges does a cube have?" the answer
"12 edges" would be marked wrong.   No marks.   Luxembourg, nul points.   
(Sorry, European joke).   The correct answer is "12".   Because to reply 
"12 edges" `implied' that a cube has 12 edges*edges.   The units are
wrong.

Andrew.