Chronology Current Month Current Thread Current Date
[Year List] [Month List (current year)] [Date Index] [Thread Index] [Thread Prev] [Thread Next] [Date Prev] [Date Next]

a different test scoring policy?



I recently had an nebulous idea on a test scoring strategy to discourage
blindly applying equations (perhaps chosen simply because the symbols
match), rather than thinking through the physics. The basic idea is to
*subtract* points when this occurs. That is, a really botched problem
could actually garner a negative score.

I have not thought at all about a more concrete implementation, nor am I
even really sure that the idea is a good one. Would the prospect of
negative scores be too frightening to students?

So I ask: Has anyone here done anything resembling this? Does it sound
like a really bad idea?

--
--James McLean
jmclean@chem.ucsd.edu
post doc
UC San Diego, Chemistry