Al Feldzamen takes a mathematicians view of how multiple choice tests work. Especially difficult ones. Turns out the harder a multiple choice exam, the lesser the accuracy of the ranking made from just counting only the correct answers!
Some of the math on that page is wrong, as pointed out in the comments on the same page, nevertheless the main argument still holds. Hmm.. SATs, GRE, GMAT, Engineering entrance exams here, Bar exams in the US..
Negative marking for wrong answers improves the accuracy a bit, but here is a very interesting scheme reported by a slashdot reader that seems to solve the problem:
Two boxes next to the answer space ask the test taker how confident (s)he is of the answer, and the scoring scheme goes:
If you ticked "confident" and you were wrong, -2
If you ticked "confident and you were right, +2
If you ticked "unsure" and you were wrong, -0
If you ticked "unsure" and you were right, +1
mighty interesting, If I ever get around to teaching a class, I might inflict this on some poor souls out there :p
Subscribe to:
Post Comments (Atom)
No comments:
Post a Comment