Ideal grading rubric for a course?

[Derek Francis]

I'd, of course, want input from people who are far more smarter and experienced than myself in academia. But I think the best and most fairest way to judge someone's grade for the course is to have a balanced variety of different means to gauge abilities.

15% Mid Term Exam
15% Non-Standardized Final Exam
Each have 100 multiple choice questions (each question is worth .1% of the final grade). Out of 15 short answer questions, exactly 10 must be selected and answered (each worth .5% of the final grade). Possibly curved up to 10% depending on the course.
10% Standardized Final Exam
115 multiple choice questions devised by the University's department heads. Automatically curved by up to 15%, to prevent a "teaching to the test" problem.
10% Quizzes
Six quizzes. Lowest score is dropped.
30% Papers
Two large papers or 7 small papers (with the weakest score dropped).
20% Homework/Lab
12 assignments total. The weakest 2 scores are dropped.

This would essentially make the course half test-based and half project-based. Among the testing, there is a variety. One under-performance wouldn't bring down one's grade, so this metric would judge by consistency. This would mean more work, but occasional mistakes are less punishing.

 

[jedishrfu]

I had a prof who did something similar in a ten week format (aka trimester) with the following added features:

1) surprise quiz every friday lowest quizzes like you did to make averaging easier

2) four tests for the trimester top three scores are used fourth test was optional so if you blow a test you have a chance to fix it.

3) final was weighted 20% to 50% of the final grade but you had to decide at the start of the test though I imagine you could decide during the test which takes some stress from the student who chose poorly.

 

[Simon Bridge]

There is no such thing as an "ideal" grading rubrick.
The person who designs the course has to balance the mentality of the students, what they want to incentivise, and how much work they want to do.

 

[Derek Francis]

3) final was weighted 20% to 50% of the final grade but you had to decide at the start of the test though I imagine you could decide during the test which takes some stress from the student who chose poorly.

thats actually a very interesting idea.

if you've struggled throughout the semester and you've eventually improved, doing well on a final that's 50% of your grade can make up for it.

at the same time, if you're more so the assignment/homework/paper student who is consistent but doesn't test very well, then the 20% won't hurt you as much.

 

[jedishrfu]

He used to grade homework but it would be selective problems and you didn't know which ones he'd grade so you had to do what was assigned.

Again some bad grades were thrown out. I think it was one or two per homework that he'd check but I don't recall.

It was Calculus 3 course and later he did the same thing in a Discrete Math course.

He was a great prof but got caught up in the publish or perish philosophy of our college. He liked teaching over publishing papers although I heard he had published some after he left our school and eventually became a dept head of math in an NYC college.

 

[Derek Francis]

Selective grading is a good way to both save time and assure people will complete every question.

Some professors only grade 3 out of 12 homework assignments in the semester. God forbid you get sick on one of those 3 days.