Re: A misapplication of probability theory in exam grading


pauldepstein@xxxxxxx wrote:
If you want the expected score of random guessers to be 0, you don't
need to fuss with silly rule changes in this way. You can just use the
normal system and then do a final subtraction of (number of
questions)/5.

Yes, the random guessers get an expected score of 0, but those who
don't answer get a negative score. I think it makes sense that they
both score 0 (one hasn't displayed more knowledge than the other).

I don't know why you use the tentative formulation "still an element"
of randomness.

Well, English is (obviously) not my native language. I mean that I
think you are caught up in the randomness, and that this system
achieves a different goal which has its merits, despite not removing
the randomness.

My whole point is that the penalty system (when used optimally by
students) has _exactly the same_ degree of randomness as the more
traditional no-penalty system.

The penalty-system was introduced with a completely dishonest and
fallacious propaganda campaign that the new system is "fairer" and
removes some randomness. It does absolutely nothing of the kind.

I think it is more fair to those who leave questions unanswered, and
agree that it does not remove randomness. I am not familiar with the
campaign.

Paul Epstein

---
J K Haugland
http://home.no.net/zamunda

.