I was poking around English Journal‘s web site to see if I could figure out when the July issue (containing my article) was coming out, and I found Alfie Kohn’s article “The Trouble with Rubrics” (PDF).

On one point, I agree with Kohn — I wish we didn’t have to assign letter grades to students, because they focus on the grade instead of on what I said was good and what needed to be improved in their writing. Grades aren’t going anywhere, however, and reformers who espouse the position that we should have no grades are generally viewed as crackpots who don’t want to be held accountable for what their students learn (or don’t want students to be held accountable). For what it’s worth, I don’t agree with that assessment, but our society is standardized-test driven.

What I don’t agree with is that Kohn sees rubrics as ineffective. Kohn argues the whole in assessment is often more than the sum of its parts. Often, students meet the criteria of rubrics, but their writing is still not good. I think when this happens, the problem is with the rubric. Rubrics should be written in such a way that they measure performance in a meaningful way. Teachers need to write rubrics that measure exactly what they seek to measure. I fully believe that many teachers don’t know how to create a good rubric. I sometimes have trouble myself. I know what I’m looking for, but how do I parse it out into different levels of achievement? Then, since I still need to give it a grade, how do I grade it?

I learned a new system this year that I will share with you. I use the Greece, NY rubrics that Jay McTighe introduced to us at a professional development session this spring. These rubrics measure five dimensions — Meaning, Development, Organization, Language, and Conventions — across six different levels of achievement. I personally think the rubrics are great. One of the problems I ran into with rubrics is math. Let’s say, for the sake of argument, that you work in a system that requires you to give numerical grades. The Greece rubrics allow for a maximum of 30 points (5 dimensions X 6 levels = 30). The problem comes when you have a student who, say, earns 25 points. We’ll say for the purposes of illustration that the student earned five 5’s — 5 X 5 = 25. However, 25/30 = 83. Wait a minute! The student scored very well on that rubric! A low B is an OK grade, but not reflective of the student’s true score. Jay McTighe showed us how to adapt rubric scores so that they are a more accurate reflection of the grade.

What you need to do is decide what the absolute lowest grade on a piece of writing should be. Most of us would not assign a score of zero to an essay. I decided that the lowest grade, if a student received five ones (the lowest level of achievement on the rubric) should be a 50. That seems fair to me — that is an F, which is indicative of the achievement of a student who received ones across the board. What this means is that the absolute zero on the assignment is not 0 points but 40 points. The difference between 100 and 40 is 60 points. As the rubric has 30 points, that means that each rubric point is worth two grade points. From there, I developed this scale:

• 30=100
• 29=98
• 28=96
• 27=94
• 26=92
• 25=90
• 24=88
• 23=86
• 22=84
• 21=82
• 20=80
• 19=78
• 18=76
• 17=74
• 16=72
• 15=70
• 14=68
• 13=66
• 12=64
• 11=62
• 10=60
• 9=58
• 8=56
• 7=54
• 6=52
• 5=50

So that means that student who earned 25 points on that rubric actually earned a 90 — much more indicative of the performance level of fives across the board.

The trick is to decide what your own absolute zero is on writing assignments and work from there. I have decided 40 is mine, so no matter how many points my rubric has, I can figure out how many grade percentage points each rubric point is worth. For instance, if I have a rubric with three dimensions and four levels of achievement, that’s a twelve point rubric. Sixty divided by 12 is five, so each rubric point is worth five points.

I think perhaps if rubrics are used in a way in which absolute zero is, indeed, zero, then perhaps Kohn has a point — they do seem punitive. But they don’t have to be used that way. I remember feeling like the proverbial light went on when Jay McTighe showed us how this could be done. Contrary to what Kohn says in his article, I think rubrics can be fair and can help students improve. I do not see them as a crutch — they are a way of demystifying something that is fairly complex — how to grade writing.