Alternative grading in algorithms
A guest post I wrote for "Grading for Growth"
Grading for Growth is a Substack newsletter about alternative grading practices written by David Clark and Robert Talbert, two math professors at Grand Valley State University. I recommend taking a look — there’s a bunch of interesting stuff (e.g., “objectivity theater,” cheating, video assessments).
Earlier this week, they published a guest post of mine; here is the link. I wrote about my first attempt at alternative grading, which seemed to go fairly well. Here are a few thoughts related to the post:
Traditionally, for a problem worth 10 points, a submission receives one of (at least) 11 possible outcomes: 0, 1, …, 10. I think having fewer possible outcomes (e.g., 2-4) makes grading faster, more consistent, and more transparent.
Most people would probably agree that students’ grades should reflect their grasp of the material at the end of the semester. If so, it’d make sense to give reassessments without penalty (but in my experience, they’re not common).
If you’re interested in the full version, please feel free to check it out!