Does Getting Rid of the ‘Zero’ Fuel Grade Inflation?
Joe Feldman, veteran educator and author of Grading for Equity, and I have been engaged in a series of conversations about “equitable grading.” In an April exchange, Joe answered some criticisms that Fordham Institute researchers Adam Tyner and Meredith Coffey had offered in a report on grading reform. Tyner and Coffey, in turn, penned a letter that ran here in May. Now, Joe shares some responses to the issues they raise, with an eye toward identifying common ground and exploring where the parties may be talking past one another.
— Rick
Rick: First off, you’ve said that you have a few thoughts on the recent RHSU letter penned by Adam Tyner and Meredith Coffey, authors of the report “Think Again: Does ‘equitable’ grading benefit students?” In the letter, they pushed back on points you’ve made with respect to equitable grading. What do you think they got wrong?
Joe: In an early part of my conversation with you, I wrote that I wasn’t opposed to penalties for late work. Tyner and Coffey allege that I’m duplicitous because I argue against lowering points for late work in my book. They miss that what I wrote was that teachers could have consequences for late work; it’s just that those penalties shouldn’t be grade-based because if they were, they would warp the accuracy of the grade as a representation of a student’s understanding. When A-level work is reduced to a B because it is a day late, and another student receives a B when it is B-level work and on time, grading becomes less accurate.
Rick: What are some consequences that teachers might use instead of grades?
Joe: Students who don’t meet a deadline could be made to stay after school to complete the work, their parent or caregiver could be contacted, or they wouldn’t be allowed to participate in a school activity until they complete the work. Many teachers who don’t penalize late work by lowering the grade help students recognize the “natural consequences” for not turning in work on time: They have to complete not just the late work but the currently assigned work, too. Plus, when students miss a deadline and points are lowered, students rush their thinking or copy another student’s assignment, or they make a calculated decision that the reduced points make the assignment not worth the effort to complete it at all, so teachers end up disincentivizing learning when a student misses a deadline. Students tell us that when deadlines are flexible and the grade isn’t lowered for late submissions, they copy less and learn more: Flexible deadlines don’t truncate student learning or send a message that deadlines are more important than learning. Teachers echo this: They see performance increases when using flexible deadlines.
Moreover, there are so many reasons why a student could miss a deadline — many of which are outside their control — such as caring for younger siblings or working. The threat of losing points is therefore meaningless: If I have to care for an older relative or need to work at my parents’ shop so we can pay rent, my teacher’s taking 5 points off when an assignment is late has no motivational power and weakens my trust that they have my best interests at heart. A final thought on deadlines: Although many will quickly say that deadlines are crucial elements of the professional world, the reality is actually quite different. Ashley Whillans, a researcher at Harvard Business School, surveyed 10,000 working adults and found that 52 percent of respondents’ task deadlines were flexible and that asking for deadline extensions reduced stress, increased performance, and was generally viewed favorably by managers.
Rick: You’ve told me that you share Tyner and Coffey’s concerns about grade inflation and motivating students to take their studies seriously. Given that, can you say more about why you think equitable grading is part of the solution here rather than part of the problem?
Joe: A decade ago, after working as a teacher, principal, and district administrator, I undertook a comprehensive examination of grading practices. I shared many of the concerns that Tyner and Coffey have expressed, including the possible unintended consequences of grade inflation and weakening motivation. As I looked at the research on grading and learned about alternative ways to grade, I thought long and hard about the usual objections: that these practices would cause grade inflation, that not including homework in the grade would result in no homework being completed, that abolishing extra credit would make it harder for students who struggle, that not penalizing late work in the grade would lead to chronic tardiness, that not including behavior-related points in the grade would be an abdication of my responsibility to teach important life skills, and other possibilities.
So, it’s not surprising and, in fact, it’s normal — and even important — that people are skeptical about some of the approaches I recommend. It’s important that we more critically examine the traditional practices and compare their goals, bases, and outcomes against those of improved, more equitable grading.
Although it is often difficult for people to be open to new ideas that contradict the beliefs they hold with certainty, given the stakes, it is nevertheless frustrating when researchers ignore evidence in favor of tropes that are simply not grounded in research. But we see it happen over and over: Teachers are skeptical that these practices will be successful — concerned that equitable grading “encourage[s] students to do the minimum,” in the words of a teacher quoted by Tyner and Coffey — but educators I work with find that with improved, more equitable grading, their classrooms become more rigorous, students become more engaged, and their professional work becomes more fulfilling. Why? Because they get to focus on teaching, assessing, and reporting what students know instead of how compliant students are. They implement grading practices that build students’ agency, accountability, and self-regulation rather than micromanage them.
Rick: One issue raised by Tyner and Coffey relates to your support for the notion that students should get no grade lower than a 50. In other words, no more zeroes. They argue that this fuels grade inflation. But you make the case that it doesn’t. Can you explain?
Joe: I’m glad that you brought that up. Many people assert that a minimum 50 percent grade on the 0–100 percent scale inflates grades by definition. After all, the thinking goes, if a student receives a 50 percent when they would otherwise have received a 0 percent, 20 percent, or anything below a 50 percent, their grade is necessarily inflated. This seems intuitive, yet it reveals a basic misunderstanding of the architecture of the 0–100 percent scale. It is a “nonlinear scale” because the range of the F is 0–60, while each other letter — D, C, B, and A — encompasses only 10 percentage points. When we use a nonlinear scale in which the range of numbers for an F is six times that of every other letter, the result is that grades are pulled downward. For example, averaging a student’s grades of a B, another B, and an F would yield a C-minus. But when we use our inherited nonlinear scale, a student who averages an 85 percent, another 85 percent, and a 0 percent receives a 57 percent. That’s an F.
To help understand how inappropriate the 0–100 scale is, let’s say the scale was reversed, so that instead of 60 percentage points to describe failure, 60 percentage points — 40–100 percent — represented the A range, with B as 40–30 percent, C as 30–20 percent, D as 20–10 percent, and F as 10–0 percent. Everyone would recognize that this would cause gross grade inflation and that allocating 60 percent of the scale to represent excellence is absurd. Assigning 60 percent of the scale to represent failure is just as absurd: It pulls grades down and therefore needs correcting. Unfortunately, this scale has been the norm in our schools for over a century. Making the minimum 50 percent makes the scale linear; therefore, it is more mathematically sound, and it prevents deflation. Interestingly, teachers find that when they use the 50–100 percent scale, they see many grades instantly increase too far, which exposes how they have been inflating their grades with “fluff” assignments. It’s like they didn’t know that their hot-air balloon had so many sandbags on it, and when they take them off, they realize that they had been using way too much propane; when the sandbags get taken off, they can significantly decrease the fuel. Once teachers make their grading scale linear, they can get rid of all of their fluff assignments, thereby increasing the grade’s accuracy.
And if we’re worried about what a minimum 50 percent does at scale, James Carifio and Theodore Carey examined 343,000 grades assigned to 11,000 students over seven years in a district that applied the minimum 50 percent to students’ end-of-quarter grades — a much more significant change than applying the 50–100 percent scale in a single classroom. They found that the 50–100 percent policy did not result in widespread “social promotion” of students. In other words, contrary to the fears of opponents of the 50–100 percent scale, Carifio and Carey found that under the 50 percent scale, students were not promoted to the next grade if they lacked the academic skills required for it.
I think it’s worth mentioning another common argument we hear: that the zero on the 0–100 scale is vital to preserve because low grades motivate students to perform better. But in fact the 0–100 scale demotivates those students who, after receiving a few zeroes, are mathematically unable to redeem themselves. The 0–100 scale isn’t just mathematically flawed; it unnecessarily punishes students. Moreover, a more linear, more mathematically sound scale can boost the learning environment of students, teachers, and schools. Carifio and Carey found that even though the change to the 50 percent minimum is “particularly challenging and even threatening” to educators who view it as a “major diminution of teacher authority,” the change to a 50–100 percent scale “actually empowers teachers and schools rather than disempowers them . . . as it creates a climate of caring, hope and support, particularly for those students whose growth and development will most probably always be an intermittent and somewhat chaotic process and path.”
I get that this is tough to understand, and even when we agree with the mathematics of it, it’s a psychological uphill climb — “How can I give students something for nothing?” we ask. That’s why I’ve found that the 0–4 is far superior to the 0–100 or the 50–100 scale, but that’s for another conversation!
Rick: OK. But, that said, it strikes me that many schools that claim to embrace equitable grading are just abolishing zeroes without this kind of careful shift to a 0–4 scale of the kind you just sketched out. Is that an issue?
Joe: Absolutely right. We find that for many districts and schools, switching from a 0–100 scale to a 0–4 scale is often a bridge too far. School and district leaders often don’t recognize how this apparently simple change of having a 50 percent minimum affects every assignment in every classroom for every student. The change, therefore, requires intensive support for teachers and a comprehensive communications plan for parents and students. And even then, it’s a constant uphill battle. We see that the minimum 50 percent is often a way station to getting to the 0–4 scale, which is much more elegant than the 0–100 scale. Besides the mathematical and statistical improvements, it matches elementary school grading and the high school GPA scale, so all students and parents are familiar with it. But shifting to the 0–4 scale requires even more significant changes to both the grading software and how teachers think about evaluating teaching and learning. Once this bridge is crossed, it causes fewer headaches and confusion than the 0–100 scale does.
This post originally appeared on Rick Hess Straight Up.
