On Sat, 28 Oct 2000 18:35:23 +0100, "haytham siala"
<[EMAIL PROTECTED]> wrote:
> Hi,
>
> I have to mark scripts based on a marking scheme thus:
>
> 10 questions of equal weighting, grade each answer from A to F. So each
> question , obviously, counts for 10%. Given a random set of grades how do I
> work out a final grade?
>
I assume that you are doing what I would call "grading papers" or
grading tests.
What I learned from a previous discussion of grading is that the
ultimate choice is yours, the teacher: how many students you want to
give A's to, or F's to?
Now, this does not apply when there are hard criteria in existence.
But if that were the case, you would not be asking about how to "work
out a final grade," "given a random set of grades."
Making up rules or procedures might be a way to draw a line for a few
marginal cases. However, it is more often an aid to self-deception,
and trying to escape responsibility. At least, that is my conclusion,
speaking as someone who has never had any of that duty.
I don't think you can do better than being arbitrary throughout:
1) Assign each Question an arbitrary number of points -- equal, if
you wish; fewer to some, if they are relatively trivial.
2) Assign letter grades to arbitrary ranges of numeric scores. If
there are gaps in the scoring, that might hint where to mark the A's
and F's (in particular). Once you draw tentative lines by
proportions, you might scan the papers near those margins to see if
you would feel better about adjusting cut-offs either up or down.
One advantage of the above is that you don't have to worry about
padding the test with easy questions, which teachers too-often feel
that they have to do, in order to assure that most people will score a
magic 70% (or whatever).
By the way, curving "up" (making 80s or lower into A's) will
probably be better received that curving "down" (making some 90s into
B's).
--
Rich Ulrich, [EMAIL PROTECTED]
http://www.pitt.edu/~wpilib/index.html
=================================================================
Instructions for joining and leaving this list and remarks about
the problem of INAPPROPRIATE MESSAGES are available at
http://jse.stat.ncsu.edu/
=================================================================
