What's Inequitable About Averaging Performance?
You’re teaching a unit on a specific skill, and during the unit, one student earns the following sequence of scores on graded assessments: 64, 70, 78, 90, 98. What grade should the student receive for that unit?
When I work with teachers and administrators on grading practices and propose that hypothetical, everyone in the room, almost without fail, pulls out their cellphone calculator and averages the scores, awarding the student an 80.
When I ask why they chose to average the scores, they often look at me with confusion. Why wouldn’t the scores be averaged?
There’s a deeply embedded belief that averaging a set of scores is the most accurate and fair way to grade — a belief bolstered and reinforced by its use in nearly all grading software. Why would we question this tried-and-true method of calculating scores?
The first problem is that the student’s average of 80 translates into a B-minus, a level the student never actually performed. We clearly are not accurately describing the student’s proficiency at the skill by giving her or him the grade of B-minus.
Also, the choice to average doesn’t capture the student’s growth and punishes a student for early challenges. It is a matter of equity. The student clearly struggled at the beginning, whether from having a weak educational background, fewer supports at home or some other factor, but by the end of the unit, the student had apparently mastered the skill or content.
Compare this to the student who had a stronger educational background or more support at home. This student might have demonstrated consistently high performance of A’s from the beginning and would therefore have a higher final grade when we averaged, even if these two students had identical performance of an A at the end of their learning.
Grades that average scores over time reflect the advantages or disadvantages of students’ circumstances and backgrounds, putting those who take longer to learn material or who start with weaker knowledge at disadvantage, and even hiding student growth.
If we grade students equitably and accurately, each of the students in the hypothetical scenario should earn an A. They started at a different place and have followed different learning paths, but each ultimately demonstrated mastery.
It should be clear how averaging performance over time would discourage students who struggle early and who are daunted by the challenge of salvaging their low initial performance. When students receive poor grades early, they may see the hill to redemption as too steep and simply give up.
Regardless of where a student started or the rate at which a student learns, grades are equitable when they reflect only whether the student, by the end of their learning, has learned the material.
— JOE FELDMAN