How to fix olympiads biggest problem. (Or at least my biggest problem)

[knottingpurple]

That's not what speed tests. If you can answer 150 simple MC questions in the time it takes the next person to answer 145, maybe you just read faster, or have better-trained test technique.

Speed on difficult questions starts to test ability rather than nonsense. Besides, difficult questions are more interesting.

Speed being a tiebreaker benefits people who have practiced taking tests over people who have studied. If you want to see who knows stuff best, very general open ended explanation questions are much better, so you can see not only who can regurgitate but who understands enough to explain well.

It's impossible to score general open-ended tests properly when most test graders are not those who have written the test or have knowledge of the event. It is also hard to grade on a holistic scale since someone might obviously have the better answer but all of them satisfy the parameters set by the answer key. Those questions are inherently flawed.

Okay, so they can't differentiate scores beyond a certain amount, so you have to have a lot of questions and very good answer keys and of course in a very understaffed tournament it might not be feasible, but they're still a lot better than almost any other type of question at identifying good competitors and definitely agree more with the spirit of Science Olympiad than just finding people who can write fast.

[knightmoves]

Okay, so they can't differentiate scores beyond a certain amount, so you have to have a lot of questions and very good answer keys and of course in a very understaffed tournament it might not be feasible, but they're still a lot better than almost any other type of question at identifying good competitors and definitely agree more with the spirit of Science Olympiad than just finding people who can write fast.

I agree, but as Raleway said, tests are usually graded by random helpers, which is why SO tends to have MC, single world answer, and numeric calculation questions - they're easy to grade, and don't require the grader to know the subject.

I'll still say that although speed at easy questions correlates entirely with test technique and writing speed rather than ability, speed at hard questions does start to correlate with ability & knowledge.