Efficient teaching: grading schemes
What is a grading scheme for our classes that promotes the best learning with minimal agony for all?
Everybody who has been teaching for a while has come upon a set of practices that works. I still don’t have a pat format, and I continue to tweak grading schemes as I gain experience, and as the content and level of the courses I teach vary. Though I don’t always do the same thing, I design my grading scheme to avoid particularly annoying or time-consuming parts of teaching, and I also make sure to include elements that are designed to improve learning. Here are key concepts that I take into account when building my grading scheme for a course:
Students that are overly stressed about the grades on their assignments are less likely to genuinely engage with the course material and will not learn deeply.
Students who are concerned that grading might be unfair will not be as interested in engaging with the course material.
Some of the students in the class might have valid reasons for needing to miss a class session. If students have worries about missing class or an assignment, regardless of the cause of these concerns, it takes a boatload of time of the instructor to deal with these concerns.
No instructor can be in a position to accurately and fairly judge the legitimacy of any excuse that a student might choose to provide.
Students learn most effectively when assessments (tests, quizzes, returned assignments) are frequent.
Students need to be able to calculate their grade in a class, and how many points remain in the semester, at any point during the semester. This requires a transparent grading scheme.
What does my grading scheme look like to deal with these principles? Here are some elements that I almost always include in my courses:
I do not take attendance, so that students don’t get overly worried if they have to miss a class.
I structure the course so that a person who chronically misses class will take a massive hit to their grade, by not having completed a number of in-class activities tied to a grade. A person who aces all of the exams, and assignment, but doesn’t do in-class assignments and quizzes, will earn a C. So, these in-class activities should constitute 20-30% of the grade.
I include some kind of frequent assessment (quizzes, homework, in-class assignments), to let students know how they’re doing. These may be graded or ungraded. Ideally every class has a very short exercise for students to size up if they’re up on the latest material.
I accept late assignments, but the moment they are late they lose 50% of their value, and it then declines by an additional 10% each week. This encourages completion on time, but still provides value to students doing the assignment later; if I did not accept late assignments then these high stakes would lead to extreme stress for some students and this would get in the way of learning. If I accepted late assignments with a minimal penalty, then too many students wouldn’t be doing their work on time and would get behind in the class.
I drop one or two of the quizzes/assignments; this can include ones not completed because a student is absent. This way, I don’t ever have to be in a position to judge whether one absence is more legitimate than another absence.
I make sure that non-exam assignments make up at least 40% of the total grade, to make sure that the exams are not high stakes.
I don’t make the final exam worth more than 35% of the total grade.
I place the first midterm exam relatively early; even with quizzes, the first exam tends to be a jolt into reality to let the students that they need to buckle down. The sooner this happens, the better.
I don’t post grades on the course management system; this keeps some students from getting any information about their grades outside the classroom, which lets them only access course content, rather than performance, when they’re not in class.
I require students to formally identify one or two partners in the course, which they may contact for a variety of questions. Students should not be asking me what they missed in a class; it’s their responsibility to find that out.
I use straight scale, such that > 90% = A; >80% = B, and so on. I only occasionally use pluses and minuses, at my discretion, and only to boost a student’s grade. I tell my students that I’d be over the moon everybody got an A, but also that that has yet to happen. This overtly encourages cooperation and group work in the course.
Grades are assigned using final scores blind with respect to the identity of the student. I sort grades from highest to lowest and evaluate the distribution. Typically, the distribution is multimodal and the grades fall out easily, and the ones on the boundaries get pluses or minuses to boost them up. I make sure that the grade a student receives is at least the minimum that they would earn under the straight scale; typically results are no different than the straight scale.
I never assign a single grade to a group assignment; I can’t see how this could not be unfair.
Once I abandoned the midterm altogether. In that class, very two weeks, we would have a short exam that took the first half of a class period. There were six or seven of these throughout the semester, and I’d drop the lowest one. The stress of midterms is gone, and students don’t stress about cramming material from long ago. There are two reasons I haven’t done this again. The first is that the one time I did it, some members of my department freaked out because it broke the mold in the department, and students in my section were happier than in other sections. (That was at a different university; now I am sure that everyone would be totally fine with it.) The second is that I haven’t had the time management skills to pull this off in future semesters. I tend to grade in big batches, and having a batch of exams every two weeks is a bit too much. I recognize that it’s better pedagogically, but I’m not sure the improvement is balanced out by the time I have to put into it.
Last semester, I offered my graduate biostatistics students the option of of a cumulative oral exam instead of a written take-home exam. Nobody took me up on that offer.
I don’t do extra credit. There’s a whole separate post about that for some point in the future. I also don’t hold out-of-class review sessions or host in-office-hour reviews with gaggles of students. And that’s also a whole ‘nother post. Students in my classes never be surprised about any exam question. That’s a third post.
What is always in your grading scheme? Do you do something overly different than me, and how do you think it affects the way that students study and learn in the long term?