Challenge a grade
Every mark on this site is open to challenge. Members can formally contest any grade and get a written adjudication, published win or lose.
Corrector right · Rubric v2.3
Every mark on this site is open to challenge. That is not a figure of speech. Correctors can formally contest any grade, and a contested grade gets a written adjudication published under the same rubric that produced it — including when the challenge wins.
Why this exists
A grading rubric that only its author can apply is not a standard, it is an opinion with a table around it. The test of whether our marks mean anything is whether someone outside this publication can pick one up, apply it to the same evidence, and reach a different answer that we then have to answer for in public.
What counts as a challenge
A challenge is a claim that a published mark is wrong under our own rubric — not that you disagree with the conclusion, but that the rubric, correctly applied, produces a different result. In practice that means one of four things:
| Ground | What you are asserting | What you must supply |
|---|---|---|
| Missed evidenceThe most common successful ground | A qualifying trial exists that we did not count. | PMID or DOI, and why it qualifies under the rubric's inclusion rules. |
| Miscounted evidence | A trial we counted should not have been, or its numbers are wrong in our table. | The specific criterion and the correct figure, with its source. |
| Misapplied rubric | The criteria were counted correctly but the grade or verdict does not follow from them. | The rubric clause you say was applied wrongly. |
| Rubric defectChanges the standard itself | The rubric produces the wrong answer here and would keep doing so. | The clause at fault and what it should say instead. |
What is not a challenge. "The dose was too low", "this worked for me", "the funding source is suspect", and "everyone knows it works" are all arguments about the world, not about the mark. They may be right and we may still be unable to act on them, because the rubric grades published trials and nothing else. If the evidence you are pointing at has not been published, the honest grade is the one we already gave it.
How it works
Adjudications are published whether or not the challenge succeeds, and a successful challenger is credited by name unless they ask otherwise. If a challenge moves a mark, that becomes a re-grade event: it is logged against the claim in the Grade Ledger and goes out as a re-grade alert.
Filing publicly is the default, because the answer belongs on the record next to the mark. A Corrector who would rather not file under their own name can send the same thing to corrections@compoundcodex.org instead — it runs on the same clocks and the adjudication is still published, with the name withheld.
If you are not a Corrector, that address is still open to you for factual errors: a misread result, a wrong figure, a citation that does not say what we claim it says. Those get corrected no matter who reports them, and the correction is posted on the breakdown. What it does not carry is a formal challenge. The named grounds, the clocks and the published adjudication are a Corrector right, and they are filed on the breakdown itself.
The standing rule
A challenge that succeeds changes the published mark, and the change is announced. A challenge that fails gets a written reason, and that reason becomes part of the record — so the next person who has the same objection can see it has already been answered, and why.
We would rather be corrected in public than be wrong in private. If the rubric turns out to be the thing that is broken, we would rather find that out from you than defend it.