Rubric — Participation and Discussion
Most participation rubrics measure volume and call it engagement. This one measures two things volume does not capture, both of which are the actual point of the course.
The first: changing your mind in response to evidence is a graded behavior. Not tolerated — rewarded, explicitly, in writing, in the syllabus students receive on day one. Students arrive believing that revising a position in front of peers is a loss, and that belief is load-bearing for an enormous amount of bad seminar behavior: the defended-to-the-death first position, the reluctance to commit to a rating at all, the student who quietly stops participating once they realize their opening claim was wrong. Grading the revision fixes more of this than any amount of encouragement.
The second: productive skepticism is not the same as reflexive contrarianism, and a course that teaches evidence appraisal manufactures the second unless it distinguishes them out loud. By week six there is usually a student who has learned that the highest-status move in the room is to doubt something. They doubt everything, at no cost, and it looks like rigor. It is the opposite: a skeptic who can be moved by evidence is doing the work; one who cannot is holding a position no evidence could touch, which is exactly what the course teaches against. Rule 4 — never downgrade with distaste — is the same principle applied to a person's temperament rather than to a molecule.
The rubric
| Criterion | Exemplary | Proficient | Developing | Insufficient |
|---|---|---|---|---|
| 1. Prepared engagement | Contributions demonstrate the reading was done and worked with — references specific studies, 🔬 callouts, or ratings by what they show rather than by conclusion. | Reading clearly done; contributions accurate and relevant. | Contributions general enough to be made without the reading. | Unprepared; contributions rely on the room to supply the content. |
| 2. Contributions are about claims | Consistently frames comments as claims with populations and endpoints. Redirects loose discussion toward what is actually being asserted. Never turns a discussion of a compound into a discussion of the people who take it. | Usually frames claims properly; occasional slippage into molecule-level talk. | Frequently discusses molecules as good or bad; needs redirecting. | Comments target people, sellers, or users rather than claims. |
| 3. Changing one's mind — the criterion this rubric exists for | Has visibly revised a position in response to evidence at least twice, said so out loud, and named what moved them. Treats revision as a result rather than a concession. Distinguishes "I was wrong" from "the evidence changed." | Has revised at least once and acknowledged it. | Adjusts position under social pressure but does not name evidence; or holds every opening position across the term without ever being moved. | Never revises. Treats challenges as attacks and responds to them as such. |
| 4. Productive skepticism vs. reflexive contrarianism | Doubt is specific and costed: names which part of the evidence is weak and what would resolve it. Applies the same standard to claims they like and claims they do not. Willing to say a claim is well supported. | Skepticism generally specific; standard mostly even across compounds. | Doubt is general ("I just don't buy it") or asymmetric — rigorous about disliked claims, relaxed about liked ones. | Contrarian by default. Objections shift when answered. No evidence in the room could satisfy them, and they cannot say what would. |
| 5. Commitment before reveal | Commits to a rating in the 📊 exercises before the book's rating is shown, in writing, and defends or revises it afterward. | Commits reliably. | Hedges until the reveal, then agrees. | Does not commit; contributes only after the answer is visible. |
| 6. Building on others | Advances the discussion: extends, tests, or steelmans a peer's reasoning before disagreeing with it. Credits where a peer moved them. | Responds to peers substantively rather than in parallel. | Contributions are serial monologues that do not connect. | Talks over others, or dominates such that others stop contributing. |
| 7. Handling anecdote and personal experience | Treats a peer's personal observation as an observation — accepted as real — and moves the discussion to the claim without correcting the person. Uses constructed examples when the structure is sensitive. | Handles others' experience respectfully; occasionally needs redirection. | Treats a peer's account as a debate target. | Dismisses, mocks, or interrogates a peer about personal use. See the grading note below — this is a floor. |
| 8. Honesty about sources | Says where a claim came from, including "I read the abstract only" or "I heard this and have not checked it." Distinguishes the three citation tiers in speech as well as in writing. | Sources contributions when asked. | Reports claims without indicating provenance. | Asserts specifics that turn out to be invented or half-remembered, without qualification. |
Scoring
Eight criteria, Exemplary 4 / Proficient 3 / Developing 2 / Insufficient 0. Maximum 32, typically scaled to a 10–15% course component.
Two structural rules:
- Criterion 3 cannot be scored Exemplary from silence. A student may only be credited for changing their mind if the revision was visible — spoken, posted, or written in a dossier revision note. Tell students this explicitly, because otherwise the quiet revisers are penalized for discretion and the loud ones are rewarded for performance.
- Insufficient on criterion 7 caps the participation grade, regardless of the other seven. This is the one behavior in the course that damages other students' ability to participate, and it should cost more than it costs the student who is merely unprepared.
Making criterion 3 gradeable
The difficulty with grading revision is that it is invisible unless you build a place for it. Four mechanisms, in ascending order of effort:
- The revision line. Every discussion post ends with one line: "Since last week, I changed my mind about _ because _" — or "nothing changed my mind this week." The second is a legitimate and fully creditable answer. Students who write it four weeks running usually notice on their own.
- The pre-commit card. In 📊 reveal exercises, students write their rating on a card before the reveal. Collect them occasionally. This makes criterion 5 gradeable and generates the raw material for criterion 3.
- The steelman turn. Once or twice a term, require each student to state the strongest version of a position they do not hold before arguing against it. Contrarianism does not survive this exercise intact; specific skepticism does.
- The end-of-term revision inventory. A half-page listing every position the student changed across the term, with the trigger for each. Pairs naturally with the dossier's drift statement, and the two should be read together — a student whose ratings all drifted upward while reporting no changes of mind has told you something useful about both documents.
Common grading edge cases
1. The quiet student who is clearly doing the work. Excellent written contributions, near-silence in the room. Do not grade the silence. Provide a written channel — discussion posts, the revision line, annotated pre-commit cards — that carries full credit, and say in the syllabus that it does. Volume is not the construct. A student who has revised three positions in writing and never spoken should be able to reach Exemplary on this rubric.
2. The confident dominator. High volume, accurate content, and criterion 6 collapsing under them — other students have stopped contributing. Score criterion 6 honestly and say what is happening in private, specifically: not "give others a chance," but "you have not built on anyone else's reasoning this term, and criterion 6 is a quarter of this grade."
3. The contrarian who is often right. Genuinely sharp, catches real problems, and never once concedes. Criterion 4's diagnostic is the single question "what would change your mind about this?" — the same question field 12 asks. A student who can answer it is a specific skeptic and should score well. A student who cannot, and who deflects the question repeatedly, is holding a position immune to evidence, and that is scored Insufficient on criterion 4 no matter how accurate their objections have been. Say the diagnostic out loud in week one so it is a known standard rather than an ambush.
4. A student discloses their own use in discussion. They may. Nobody may be asked to. Your handling is graded by the room whether or not it is graded on paper: accept the observation, thank them, move to the claim. Then score other students on criterion 7 for how they respond — and be prepared to intervene immediately if the room turns the disclosure into a case study. The student who disclosed is not scored differently for having disclosed, in either direction.
5. A student changes their mind toward a position you think is wrong. Full credit. Criterion 3 grades the process, not the destination, and the standard from the dossier rubric applies here too: a well-reasoned ⚠️ that the instructor disagrees with outscores a ✅ that happens to match the book. A student who moved because of evidence and landed somewhere you would not has done the graded thing.
6. Revision that is actually capitulation. A student who changes position every time they are challenged, without ever naming what moved them, is not scoring Exemplary on criterion 3 — they are scoring Developing, and the feedback is precise: you revised, but you did not say what evidence revised you, and I could not tell the difference between being persuaded and being outlasted.
7. Dosing, sourcing, or protocol talk in discussion. Course policy covers speech as well as submissions. Redirect in the moment — briefly, without embarrassment, and without heat, because heat is what teaches concealment. "We keep protocols out of this room; the claim underneath it is fair game, so what is the claim?" is usually enough, and it converts an out-of-bounds contribution into an in-bounds one, which is a better outcome than shutting the student down.
Related: Chapter 5 · Chapter 6 · Chapter 39 (the conversation; the same tone applies in a seminar) · Appendix C §C.6 (drift, and measuring your own bias) · How to Teach This Book (classroom management) · Where Students Get Stuck (struggles 9 and 10) · Evidence Dossier Rubric · Claim Evaluation Rubric