Rough Grading Rubric
Spring 2018
In an attempt to make grading for this class a little more consistent and
transparent, I am publishing a rough grading rubric to be used on exams and
homeworks. This will not be applied exactly, as different problems have difficul-
ties in different places, and we will often want to keep scores at round numbers
when possible. Hence this is intended more as a rough guideline for the types
of things that will get credit.
Firstly, your solution will be sorted into one of a few categories:
• Relevant Observations: Made basic, relevant observations that are not
useful towards finding a solution: ≤ 15% credit.
• Trivial Algorithm: Correct but slow algorithm, each to find: ≤ 25% credit.
• Partial Solution: Made observations that are a substantial part of a full
solution: credit varies based on how much of the full solution was obtained.
• Unoptimized Algorithm: Correct algorithm, slower than asked for, but
non-trivial to find: credit varies.
• Full Solution: Full solution with a correct proof: 100% credit.
Errors: points can be removed in you have errors in your solution
• Minor errors: Computational mistakes, easily fixed: ≈ −5% credit each.
• Major errors: Missed some case, or proof has some small conceptual hole,
not too difficult to fix once pointed out: ≈ −20% each.
• Substantial error: Major gap in argument or wrong approach: Variable,
can downgrade solution category.
1