Most rubrics grade what a student said. This one grades what would make the student wrong. That small shift changes the classroom.
The shift, in one sentence
A claim without a falsifier is a wish. A claim with a falsifier is an argument you can share with a room. So the rubric awards points for the sentence a student writes after their claim: "This would be wrong if..."
Where the idea comes from
The habit is older than any of us. Karl Popper argued that a theory earns its keep by forbidding something (Class E). If nothing could count against it, it is telling us nothing new. Working scientists internalize this: they design experiments that could embarrass their own hypothesis. The rubric below is a small pedagogical translation of that stance, sized for a class period.
It also sits inside the active-inference framing we teach at EducateWright: a learner is a model of the world, and models update when predictions meet evidence (Class E, Parr, Pezzulo, and Friston, 2022). A falsifier is just the student naming, in advance, the observation that would force an update. Nothing mystical. Just honest bookkeeping.
The rubric (4 rows, 16 points)
You can staple this to a lab report, a history essay, or a math justification. It travels.
Row 1: Claim clarity (0 to 4). Is the claim a single sentence a stranger could repeat? Vague claims cannot be falsified, so vagueness costs points here, not later.
Row 2: Evidence class tag (0 to 4). Did the student mark the claim with an evidence class: A empirical-in-session, B code or inspection, C configuration or integration, E expert citation, F falsifier present, or U unverified? An untagged claim caps at 2. A correctly tagged U (honest "I do not know yet") gets full marks. Honesty is graded.
Row 3: Falsifier sentence (0 to 6). The heart of the rubric. Full credit requires a concrete, observable condition that would make the claim wrong, written in the student's own words. "If we ran the pendulum at a longer length and the period did not increase, this claim would be wrong" gets 6. "If it turned out to be false" gets 0. We are grading specificity, not eloquence.
Row 4: Update plan (0 to 2). One sentence: "If the falsifier hits, here is what I would believe instead." This closes the loop and prevents the falsifier from being decorative.
Sixteen points, one page, gradable in about ninety seconds per paper once you have run it twice.
Why the update-plan row matters
Without Row 4, students learn to write falsifiers as a ritual and ignore them. With Row 4, the falsifier becomes load-bearing: the student has already rehearsed the belief revision, so when contrary evidence arrives, the update is cheap. This is the classroom analog of a system that is designed to change its mind cleanly (Class C, based on how we configure our own gates in production).
Common failure modes, and how the rubric catches them
The unfalsifiable reframe. "My claim is that this poem is beautiful." Row 1 fails: the claim as written cannot lose. Coach the student toward a claim that can: "My claim is that the poem uses more monosyllables in stanza two than stanza one." Now Row 3 becomes tractable, and the aesthetic conversation gets richer, not thinner.
The circular falsifier. "This would be wrong if the claim were false." Row 3 fails. Push for a concrete observation.
The tag-and-run. Student writes "Class A" on a claim they never actually observed. Row 2 fails; a wrong tag is worse than no tag. This is where the rubric teaches the vocabulary of evidence classes as classroom vocabulary more effectively than any lecture.
What changes in the room after two weeks
Students start pre-writing their falsifiers before writing their claims, because they have learned that a claim they cannot falsify is a claim they cannot defend. Discussions get quieter and more honest. Disagreements become useful: two students with different claims can now compare falsifiers and know exactly what evidence would settle it. That is the whole point.
Where this rubric fits in your curriculum
- Pair it with Gate Design for Middle School so the rubric grades the same falsifier the gate later checks.
- In science, run it alongside Gate Design for High School Science for lab-report continuity.
- Come to a workshop and we will grade a stack of your own student work with you, live, and tune the rubric to your subject.