Grading and XP
© 22 May 2013 Luther Tychonievich
Licensed under Creative Commons: CC BY-NC-ND 3.0
other posts


What does a point-based grading scheme suggest about the teacher’s model of learning?


A week or two ago a colleague shared an idea to use “‍experience points‍” to track grades in a game design class he is teaching. This thought brought to mind two ideas. First was “‍most of my classes have worked like that: each assignment you do gives you ‘‍points‍’ toward an A.‍” The second was “‍what a poor model of learning!‍”

Point-based grading rewards doing work and knowing things on schedule. But learning does not happen on a schedule. I suppose that’s because people are individuals and the labor of developing new mental habits and abilities varies based on what particular skills an individuals’ mind has already gained.

Point-based grading is appealing in several ways. It is easy as a teacher: simple to design, unambiguous in scope, with effort distributed across the term. Incorporate incentives for almost any desirable behavior is straight-forward. And if the goal is to grade obedience or diligence rather than knowledge or skill acquired it also makes a good deal of sense.

There’s another way that point-based grading might make sense. Despite a complete dearth of evidence, the view persists that intelligence and aptitude is somehow inherent, not developed; I’ve written of this “‍before‍”. If you ascribe to this fallacy then an “‍I teach it, I measure if you ‘‍got it‍’, and then I move on‍” model of teaching and evaluating learning is natural. But usually courses are built to teach ideas that build off of, and reinforce, other ideas in the course and it is quite possible that someone who didn’t succeed at learning a concept when it was first presented will end the course with a better knowledge of it than someone who did.

An attractive alternative to the point-based model allows students to re-examine themselves as often as they wish. When grades are due each receives a grade based on the skills they mastered by that point. The problem, though, is that, hard as it is to test knowledge once, it is far harder to test it reliably over and over again.

I’ve written before about not knowing what ought to be graded in the ideal world. Practically, though, it is the difficulty of how, not the ideal of what, that drives my exam development.

Looking for comments…

Loading user comment form…