Zeros and the 50 Floor — Teaching

Start here

The math both sides start from

Everyone in this debate agrees on the arithmetic. They disagree about what students do once they understand it.

The 100-point scale as most grade books use it

60 points of F, about 10 points for every other letter

F

D

C

B

A

05060708090100

The same scale with a 50 floor

Every letter gets roughly 10 points; nothing below 50 is recorded

not used

F

D

C

B

A

05060708090100

The example the "change" side uses

A B student misses one assignment

Two 85s, then a missing assignment	Average
Recorded as a 0	57 (F)
Recorded as a 50	73 (C)
More 85s needed to get back to an 80, after a 0	14
More 85s needed to get back to an 80, after a 50	4

One zero can outweigh a semester of B work. This is Tamony's "kill grade" and Wormeli's "six or seven perfect 100s to reach a D."

The example the "concerns" side uses

An A student stops turning things in

Three 95s, then ten missing assignments	Average
Missing work recorded as 0s	22 (F)
Missing work recorded as 50s	60 (D)
Share of assignments actually completed	23%

With a 50 floor, a student who does a quarter of the work can still pass. This is Rablin's example: three 95s, skip the next ten, and still pass.

Two sides, three pieces each

The case for the floor, and the case against it

Each card gives the argument in a few sentences and the claims you're most likely to hear repeated. Read the originals before quoting them.

Make the change Raise concerns

Make the change

Replace the zero with a 50

The scale is the problem. A 50 is still an F; it just stops one missing assignment from erasing a semester.

Washington Post Answer Sheet, November 14, 2023

Rethinking the way teachers assign student grades

Guest essay by Rick Wormeli, in Valerie Strauss's column. Wormeli is a former National Board Certified teacher and the author of Fair Isn't Always Equal.

Wormeli's case is that a grade reports learning, not compensation, so "something for nothing" is the wrong frame. A 50 is a placeholder that says no proficiency was shown and no credit was earned. What a 0 does is mathematical: on a 100-point scale it can take six or seven perfect scores in a row to climb back to a D. He argues punitive grading has no research behind it as a way to build responsibility, and that missing work is often a signal of hardship rather than character.

Claims you'll hear

A 50 does not mean "half the learning." It means failure, recorded on a scale where every letter gets the same room.
If you must average on a 100-point scale, turn zeros into 50s, or use the mode instead of the mean.
A recoverable grade teaches responsibility; an unrecoverable one teaches students to quit.

Read at the Washington Post

Paywalled. The full essay is also posted at the National Education Policy Center.

Edutopia March 17, 2023

Why the 100-Point Grading Scale Is a Stacked Deck

Youki Terada, who covers education research for Edutopia.

A history of the scale: Harvard adopted 100-point grading in 1837 as a ranking tool centered on 50, and it was later merged with letter grades in a way that left 60 percent of the scale to failure. Terada argues a zero produces helplessness rather than correction, especially early in a term, and that the "real world" argument for hard deadlines is weaker than it sounds. He lays out alternatives: minimum grading, dropping the lowest score, allowing make-up work, 4-point standards-based scales, and delaying grades in favor of feedback.

Claims you'll hear

The 100-point scale was never designed to measure learning; it was designed to rank.
Carifio and Carey's seven-year study found minimum grading went with higher state exam scores and graduation rates, not grade inflation.
A 2022 workplace study found more than half of real deadlines are flexible, so strict late penalties don't mirror adult life.

Read at Edutopia

Edutopia October 6, 2021

The Case Against Zeros in Grading

Alexis Tamony, a math teacher in Sonoma, California, then in her 19th year.

A teacher's account of changing her own practice. Tamony calls the zero a "K grade": if each letter got ten points, a 0 would sit four letters below F. She walks through the B student who misses one assignment (57 with a zero, 73 with a 50) and reframes the floor as a conversion of the 0 to 4 scale onto 50 to 100. She is candid about the pushback, mostly from parents who see a "7/14" next to a missing assignment, and argues that defending the practice to colleagues was the best professional development she has done.

Claims you'll hear

A student can have fifteen 85s and one zero and still end with a C.
A 50 floor makes each grade band worth the same, so failure isn't weighted more than success.
The online grade book, not the policy, is what confuses families; communicate the rationale.

Read at Edutopia

Raise concerns

The floor fixes the math, not the problem

Students respond to the new arithmetic too. Take away the zero without replacing what it did, and effort drops.

Edutopia March 28, 2025

Getting Rid of Zeros Won't Fix the Grade Book

Tyler Rablin, a teacher who has written a book on grading and motivation.

Rablin grants the case for the floor, then describes what teachers in 50-floor schools report: once students work out the math, some stop turning things in. A student with three 95s can skip the next ten assignments and still pass. His answer is not to restore the zero but to replace it with a real accountability lever. He used a late-work contract: an incomplete, two required tutoring sessions, and a consequence matched to the behavior behind the missing work, with no lasting grade penalty once completed. He argues the deeper problem is averaging over time.

Claims you'll hear

Without a new accountability measure, a 50 floor lowers engagement and eventually learning.
Any system that removes a lever has to put another one in its place, ideally one that builds habits.
The 50 treats a symptom; the disease is averaging.

Read at Edutopia

School Administrator AASA, December 2024

Unintended Effects of 'Equitable' Grading

Adam Tyner, national research director at the Thomas B. Fordham Institute.

Written for superintendents. Tyner uses Fairfax County, Virginia, which adopted a package of reforms in 2022 (a 50 floor, no late penalties, retakes, limits on homework) as his case study, drawing on teacher accounts. A veteran high school teacher estimates that at least 30 percent of her students visibly reduced their effort once the rules changed. His argument is that reforms meant to level the field instead lower the expectations placed on the students they were meant to help, and that districts adopt them faster than the evidence warrants.

Claims you'll hear

Post-pandemic interest in "equitable grading" has run ahead of evidence that it works.
Teachers in reform districts describe too much leniency and less student effort.
Lower expectations are not equity, even when the intent is.

Read at AASA

Educational Leadership ASCD, October 2024

The Unwinnable Battle Over Minimum Grades

Thomas R. Guskey, Douglas Fisher, and Nancy Frey. Guskey is professor emeritus at the University of Kentucky and a longtime grading researcher.

The concern here is different: minimum grades are a fight nobody wins because they treat a symptom. Teachers resist recording credit that wasn't earned, families see social promotion, some students game the floor, and even with it in place a teacher is still sorting work into 51 flavors of F. The authors' fix is to leave the 100-point scale behind: grade in four to seven categories, stop averaging, base the grade on the most current evidence of what a student can do, and write down what a grade is supposed to mean.

Claims you'll hear

The percentage scale creates an illusion of precision and makes teacher agreement harder.
Averaging punishes risk-taking because everything counts forever.
Adopt a floor and you have changed the arithmetic; you have not changed what the grade means.

Read at ASCD

Peer-reviewed and near it

What the research actually says

Each entry names the design so you can weigh it. The chip shows which side the finding tends to support, not which side the authors are on.

The short version. There is no meta-analysis of minimum grading, and no peer-reviewed study that puts a 50-floor school and a zero school side by side with a real comparison design. The one direct study is a single school that already used the floor. The strongest studies in the area test something adjacent: what happens to effort when grades get easier to earn, or when a teacher grades more strictly.

Read together, the evidence supports two claims at once. Zeros do mathematical damage that a floor repairs, and easier grading does reduce effort among some students unless something else holds them to the work.

Direct studies of the 50 floor

All from one research team and one Massachusetts high school, so treat them as a case, not a verdict.

Supports the floor

Carey & Carifio (2012) Educational Researcher 41(6), 201–208

Seven years of records (2004–2010) from one large urban high school that already used a 50 floor: 343,425 quarterly grades from 851 teachers. About 8.5% of grade sets began with a minimum grade, and only 0.3% of all grade sets ended in a passing course grade because of it. No evidence of grade inflation or social promotion. This is the study Rablin means when he says "there's research claiming" the floor doesn't inflate grades.

One school, no comparison school, observational. The authors are open advocates of the policy.

Journal page ERIC record
Theory only

Carifio & Carey (2009) The High School Journal 93(1), 23–37

The argument for minimum grading built from motivation theory (locus of control, self-efficacy). Concedes the benefits are largely theoretical and calls for empirical tests; suggests the students most likely to benefit are those in high-risk courses.

ERIC record
Supports the floor

Carifio & Carey (2013, 2015) follow-up papers on the same data

Re-analyses of the same school's records. Students who received minimum grades did as well or better on the state exam once GPA was controlled; about 142 students a year passed a course they would otherwise have failed. Terada's article cites this as "29,000 students"; in the paper, 29,187 is the number of quarterly grade sets that began with a minimum grade.

Same single-school data as the 2012 study.

2013 point-counterpoint (PDF) 2015 paper (PDF)
Supports the floor

Morris & Johnson (2024) Education Policy Analysis Archives 32

A simulation, not a trial: what a statewide ninth-grade minimum grade would have done for the Arkansas class that started high school in 2015–16. Estimates it would touch 1 to 3% of students and convert roughly 10% of failing grades to passing, at no added cost. Models a policy; does not test one.

Journal page (open access)

Causal studies of grading leniency and standards

These don't test a 50 floor. They test the mechanism the concerns side is worried about: what students do when grades get easier or harder to earn. The designs are stronger.

Raises concern

Bowden, Rodriguez & Weingarten (forthcoming) American Economic Journal: Economic Policy

"The Unintended Consequences of Academic Leniency." When North Carolina moved every high school from a 7-point to a 10-point scale (an A dropped from 93 to 90, passing from 70 to 60), the authors compared the last cohort under the old scale with the first under the new one. GPAs rose about 0.27 points, end-of-course test scores stayed flat, lower-performing students' attendance fell, and achievement gaps were wider by the ACT. The cleanest evidence that easier grades reduce effort among the students the change is meant to help.

A scale change, not a floor. Statewide, cohort comparison.

Journal page Working paper (free)
Raises concern

Gershenson, Holt & Tyner (2024) Contemporary Economic Policy 42(2), 305–318

A decade of statewide Algebra I records (2006–2016) linking students to teachers. Students assigned to teachers with tougher grading standards learned more, in that course and in later math, with slightly fewer absences. Effects were similar across racial and income groups. Tyner is the author of the AASA piece above.

Journal page Working paper (free)
Raises concern

Figlio & Lucas (2004) Journal of Public Economics 88(9–10), 1815–1834

Grades 3 to 5 in Alachua County, Florida, 1995–1999. Teachers with higher grading standards produced larger test-score gains in reading and math and fewer discipline problems. Gains depended on classroom mix; the first study to measure standards at the teacher level.

Journal page NBER working paper
Cuts both ways

Betts & Grogger (2003) Economics of Education Review 22(4), 343–352

National High School and Beyond data. Higher school-level grading standards raised test scores, mostly at the top, but did nothing for educational attainment and were associated with lower graduation rates for Black and Hispanic students. The concerns side's mechanism and the change side's worry, in one dataset.

Abstract

Reviews and syntheses

Background

Brookhart, Guskey et al. (2016) Review of Educational Research 86(4), 803–848

"A Century of Grading Research." The closest thing to a comprehensive synthesis. It doesn't address minimum grades, but it establishes why the debate exists: grades consistently measure a mix of achievement and non-cognitive factors like effort and compliance.

ERIC record Author copy (PDF)
States the gap

Hutt (2025) AEFP Live Handbook, "Grading Policies"

A research synthesis for policymakers from the Association for Education Finance and Policy. Walks through the studies above and says plainly that the "no zeroes" push rests on a true observation about the scale but that there is no direct evidence on whether the policy works.

Read the chapter
Raises concern

Coffey & Tyner (2024) Thomas B. Fordham Institute, "Think Again"

A think-tank review, not peer-reviewed, and Fordham has a position. Useful because it gathers the standards research in one place and admits it has no direct evidence on 50-percent floors either; its case against them is reasoned from the leniency studies.

Read the report

The only head-to-head comparisons are dissertations

Not peer-reviewed, both confounded, and they point in opposite directions. Included because they are the only studies that compare a floor school with a zero school.

Raises concern

Dennis (2018) Ed.D. dissertation, Walden University

Followed 1,396 high school students from two feeder middle schools, one with a no-zero policy and one without. Students from the no-zero middle school earned somewhat lower high school grades in math, science, and English (small to medium effects).

Dissertation record
Supports the floor

Mariani (2023) Ed.D. dissertation, Liberty University

One suburban Pennsylvania high school (about 1,250 students) before and after adopting a 50 floor. Final grades and biology and literature Keystone scores rose; algebra did not change significantly.

Dissertation record

Common ground

Where the two sides agree, and where they don't

Both sides say the scale is skewed

None of the six defends the zero's arithmetic. Tamony, Terada, and Wormeli lead with it; Rablin concedes it in his opening; Guskey, Fisher, and Frey want the percentage scale gone altogether.

Both sides say averaging is the deeper problem

Rablin ends with it, Guskey builds his whole argument on it, and Wormeli's "use the mode" is a version of the same point. A floor is a patch for averaging; several authors on both sides would rather stop averaging.

They disagree about what students do next

The change side expects a recoverable grade to keep students working. The concerns side expects some students to do the new math and work less, unless something other than the grade holds them to it.

They disagree about what replaces the zero

For Tamony and Wormeli, communication with families. For Rablin, a contract and an intervention period keyed to missing work. For Guskey, Fisher, and Frey, a different scale entirely. Tyner doubts the reforms deliver what they promise.

For the conversation

Questions worth bringing to the table

In our grade books, what is a missing assignment supposed to communicate: no evidence yet, or a refusal? Do our numbers say that?
If we adopted a floor, what would replace the zero as the reason to turn work in? Who would run it, and when?
Which of our students are the ones a floor would actually touch? The Massachusetts data says mostly working-class boys; is that true here?
Is the argument really about 0 versus 50, or about whether we keep averaging on a 100-point scale at all?