No Zero Grading And The Mystery of Assessment

It's a story that periodically resurfaces. Just last week, there was a recycling of the tale of Diane Tirado, the Florida teacher who was allegedly fired for giving students a zero on homework. 

There are plenty of lessons to be learned from this story. One is the dangers of teaching in a state with few job protections and deliberately weakened unions for teachers. Another is to beware of these sorts of viral stories, because maybe there were other reasons she was let go and maybe the school doesn't actually have a no zeros policy.

But even though the story is five years old, it still has legs because no-zero grading is reliably click-generating and button-pushing. For some folks it just rings the fuzzy-headed liberal bell. Your uncle who bitches about participation trophies also hates no-zero grading. That's not how life works. It's coddling students. And, as this conservative writer puts it,

It’s part of being nice and progressive, considerate of students’ feelings and respectful of their egos.

I know most of the arguments from the years we debated a no-zero policy in my district. We had switched from a letter grade system for nine weeks grades to a 100-point scale, and shortly after, the district created a policy that no student could receive a grade lower than 50% in the first grading periods of a course. 

There was nothing nice or progressive about it. It was practical matter of teacher preservation. 

Under a letter grade system, with averaging math based on a four point scale (A=4, B=3, C=2, D=1 and F=0), an F was not a grading catastrophe. But with a 100 point scale...well, imagine a worst case scenario. With 70% the lowest passing score (as it was for many years), a student needs to hit 280 total for the year. If the student pulls a 0 in the first nine weeks, that student needs to hit the low nineties in the remaining three quarters.

Again, the no-zero policy was not about the student's tender feelings or vulnerable ego. It was about the problem of spending 135 days with a teenager in your classroom who knows he cannot possibly pass your class, that his failure is already written in stone, and so has A) no reason to try and B) nothing to lose. A no-zero policy is not doing the student a favor; it is giving the classroom teacher one more chance to hold onto one more piece of leverage for just a little bit longer.

A discussion about no-zero grading opens the door to a discussion of grading itself, which is a difficult topic mostly because it involves doing the impossible--assigning a clear and specific numeric value to the learning that may or may not have occurred inside a student's head.

While attempting to do that impossible feat, we layer on issues of number crunching. One of the things my district didn't like about the four point letter-to-number scale is that such a scale effectively awards students the lowest possible score for the grade (if 3.0 -3.9 is a B, but a B is worth 3.0, then the student is always just scraping by). Should we weight the difficulty of the course? Should we weight for times of year--isn't learning demonstrated in May more important than learning demonstrated in September? But if we weight, how do we build that into numbers? Do we get into the mathematical business of curving, even though drawing the lines on the curve is an arbitrary task often motivated by non-educational concerns.

And if this grading stuff is hopelessly messy, can we just not? Go back to the idea of portfolios, because maybe now that we have digital capabilities portfolios won't be such a giant unmanageable mess of... nah, they'd probably still be a giant mess. Maybe competency based stuff, with a giant checklist of bare minimum demonstrations of various "skills" except that bare minimums are problematic and not everything students learn is a skill and if you wander too far from a traditional approach, expect a massive wave of parents asking "What the hell does this even mean? Can you just tell me how my kid is doing??"

What are grades supposed to communicate? Is there an absolute amount of stuff that students should have mastered by a certain grade, and who decides, and do you just use data about what the average student has mastered in which case A) you're back to grading on the curve and B) roughly 49% of students will ne below average, always. 

Should a grade tell what a parent how a student compares to other students, or to some objective standard? Should a grade communicate something useful to future employers, or colleges, or to the student's future teachers, or to the state itself? 

Is it possible to come up with a system that does all these things at once, or shall we just use a tool designed for one purpose for other purposes as well? Can we use a hammer to put in screws? And can we keep from polluting the data with other information that weighs factors like compliance and attendance, neither of which are trues measures of what the student has learned.

And all of this because at the heart of assessment, we are trying to know the unknowable--what is going on in another human's head. We can claim that performance tasks fill the bill, but that will always include the extra layer of a student's ability to perform. Some students are great at performing despite not really knowing. Some students who really know can't perform. Some students just don't want to perform. And what about all the stuff that just kind of lies dormant upstairs until a light bulb goes on five, ten, twenty years from now. 

I'm not saying throw up our hands and claim that it's impossible, so don't even try. I am saying that any time someone starts talking about grades or test scores or assessments as if they are solid, absolutely reliable numbers that precisely represent reality, they are shoveling baloney. Maybe it's wishful thinking baloney, or maybe it's self-deluded baloney, or maybe it's let-me-sell-you-some-snake-oil baloney. But it's still baloney. 

All I'm asking is that we talk about grades and assessment for what they are--our best attempt at getting an approximate read on what is going on inside the head of a particular young human being. No assessment or reporting system is so finely, perfectly attuned that to alter a piece of it would be a crime against its shiny perfection. All grading systems contain a not-inconsiderable amount of junk. 

That's okay--as long as we remember that we aren't dealing with perfect systems. But it is very human to make shit up and then pretend that that shit descended from the heavens on a silver platter carried by the hand of God. 

We do the best we can with what we have and what we know, and we accept that we can, and should, always try to do better. That's a fundamental rule for being human in the world, and it covers assessment of our fellow humans as well.