In article <8tg0i6$k85$[EMAIL PROTECTED]>, <[EMAIL PROTECTED]> wrote:
> ... I am grading an HTML coursework. I have
>to look at their web pages and mark them against ten requirements given
>me as criteria, and grade each from "A" to "F". Unfortunately the
>schema is flawed because for example: "is xxx link working?" is best
>treated as boolean and is not really gradable as a "B" for example. I
>suppose an "A" if working and an "F" else - will suffice.
>
>If the marking scheme were better thought out I could perhaps give
>marks in accordance with some plan and then by applying the
>University's scale award a grade. Unfortunately as a lowly research
>student / teaching assistant, I have had no input into this and I
>suspect that there will be a different grading scheme for each lecturer
>and standards will vary wildly.
>
>I expected this to be a fairly common problem and hoped that a
>standard "formula" would already exist for it...
Your only sensible course of action is to talk to whoever is actually
in charge of this course and determine what they want you to do.
Otherwise, as you say, the marks will vary from section to section in
an unfair manner.
If there is no one else in charge, then you should do whatever you
think is most fair, including perhaps modifying the letter grade
scheme, unless some committment has been made in that regard to the
students.
Statistics will not tell you what is fair.
SOMEBODY should be responsible for ensuring that the grades are fair.
If that person is you, then do it. If it's not you, then ask the
person whose responsibility it is. If someone is supposed to be
responsible, but is not doing their job, then you have a difficult
problem that you'll have to figure out how to deal with somehow.
Radford Neal
=================================================================
Instructions for joining and leaving this list and remarks about
the problem of INAPPROPRIATE MESSAGES are available at
http://jse.stat.ncsu.edu/
=================================================================
