This page describes my approach to exams, how they are graded, and how I handle requests for re-grades.
First, an important point - for most of you time with me (homeworks, lectures, quizzes, etc...), the primary objective is learning. Thus, I give you full and extensive solutions, the TAs and I will happily discuss your solutions with you, and we'll generally be as transparent and open as possible. Sometimes, this gets in the way of fairness - for example, students who are more pushy about getting their homeworks re-graded might get a few extra points. But that's a price I'm willing to pay for the sake of creating a more effective and meaningful learning experience.
For the final exam, my priorities are reversed. I'd love you to learn something from your experience taking the final, but my overwhelming priority is to assess you, and, crucially, to do it fairly. This is the fundamental driving principle behind all my exam policies.
Now, to specifics:
- I will never share either the exam paper, or its solutions, with anyone. These remain confidential so that I can re-use questions over the years while ensuring the exam stays fair. Why re-use questions? Simply because the newer an exam question is, the more likely it is the phrasing might be confusing, thus making the exam less fair. By combining old and new questions in each exam, I ensure the bulk of the exam is tried-and-tested.
- If the exam was a multiple-choice or numerical-answer-only exam, I will not entertain any requests for re-grades or go through any exam.
- If the exam is not multiple choice, rest assured I will have taken a number of extensive steps to ensure the grading was fair. To name just a few:
- Every question is graded by a single TA, to ensure points are deducted consistently throughout.
- I ask every TA to first look through a number of copies before starting to grade, to develop a consistent grading scheme.
- I randomly check a sample of exams at the end of grading, and re-grade them from start to finish myself to ensure the TA's choices are reasonable.
- I do a perturbation analysis on every grade to see what the effect of your getting 3-5 extra points on your exam would have been. If this would have made a difference to your final grade, I will ask the TAs to look at your exam in more detail to ensure it was graded fairly and consistently.
- If you nevertheless feel like you did much worse than you expected and that there might have been a major error in your grade, you may reach us to us and ask for a re-grade. The procedure is as follows
- Request a re-grade by emailing the class email address. When you do this, please mention you read this page in full.
- The TAs will regrade your entire exam, from start to finish. This could result in your grade going up or going down.
- The TAs will reach out and let you know what the result of the re-grade was.
- If you still feel your grade does not reflect your performance, the TAs will hold office hours during which you can go through your exam with the TAs. If, as a result of this meeting, the TAs feel you should be granted additional points, they will refer the entire final to me. I will then re-grade your entire final again in light of this information, as a result of which your grade may go up, or down. This is to ensure we don't just award extra points to those who shout loudest.
The primary complaint I get as a result of these policies is "but I don't want extra points, I just want to see where I went wrong." In answer to this question, remember: much as it pains me to say this, my priority in the final is not to ensure you understand where you went wrong, but instead to make sure you are assessed fairly. This is a necessarily evil of the way the academic world works, and the fact I have to grade classes on a reasonable curve. From past experience, allowing people to review their finals freely results in everyone who makes an appointment getting a few extra points back, which is a textbook example of an unfair advantage for some people (perhaps those with the time to make that appointment). Thus, after much soul-searching, I have converged to the scheme described on this page for optimal fairness.
I, of course, remain open to discuss this policy, and available to answer any questions you might have. I'm absolutely open to be convinced that there is a better, or fairer, policy I should adopt.