[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Various things

On 20050721T213458-0500, Peter Samuelson wrote:
> People keep complaining about the Essay Test from Hell, but I guess I
> still don't quite see the problem.  Who is it that doesn't like it?  Is
> it AMs that feel it's too much work on top of their other duties to get
> to know an applicant's skill level?  Is it the competent applicants who
> feel it's beneath them?  Or is it incompetent applicants who feel
> intimidated by all the hard concepts to look up?

The Essay Test from Hell is exactly like a traditional unviersity exam,
at least as they are performed around here: 1) it's a lot of work, 2) it
is stressful for the student, 3) it seems to measure really well what
the student/nm has learnt and 4) it actually does not measure anything
of the sort.  What people learn from traditional exams is not the
subject matter but how to pass traditional exams.  (Most people do also
learn the subject matter, but that's mostly in spite of, not because of,
the tests - they realize that they will need to know the stuff after the
test, as well.)

My day job is half teaching, and that responsibility also carries the
need to prepare exams for my students and to grade them.  While I am not
in any sense a formally qualified teacher (if you must ask, do it
privately), I have done some reading on didactics, and especially
university-level exam theory and practice.  All my sources agree that
the traditional exam is a very bad method of conducting exams.  The best
method is to give the student a real task that they should perform (eg.
programming students should be made to create a program that solves a
real problem for a real user), and grading should be based on observing
both the process and the end result (and the user's satisfaction!); when
that is impractical, the students should be given tasks that simulate
real tasks (like programming a solution to a realistic but still
invented problem).  Only when that is also infeasible, should
questioning be used, and it should be realized that it is a poor
substitute for the better methods.

Antti-Juhani Kaijanaho, Debian developer 


Attachment: signature.asc
Description: Digital signature

Reply to: