Checking
Criteria written first
You have to judge a piece of work, yours or someone else’s, and “it’s good” or “it doesn’t work” is all that comes to you.
How it goes
do it yourself- Write five criteria BEFORE showing the work, each phrased as a closed question: “does it cite a source for every figure?”
- Have the work marked criterion by criterion, with the exact quotation justifying each mark.
- Do not argue with the marks. Argue with the quotations: a mark without a quotation is an opinion dressed up as a measurement.
The instruction, word for word
copy it, then correct itHere is my grid. I will give you the text to mark AFTERWARDS. 1. Is every figure given accompanied by its source? 2. Does the recommendation fit in one findable sentence? 3. Are the predictable objections addressed, or avoided? 4. Can a reader new to the subject act on it after reading? 5. Is there a claim presented as fact that cannot be checked? For each criterion: yes / partly / no, THEN the exact quotation from the text that justifies your answer. If you cannot find a quotation, the answer is "no" and you say so. Give no overall mark. Do not rephrase my criteria.
- The grid given first
- The whole exercise rests here. Criteria written after reading conform to the text read and stop measuring anything.
- Closed questions
- “Is it clear?” cannot be checked. “Does the recommendation fit in one findable sentence?” is checked by looking for it.
- “If you cannot find a quotation”
- The rule that blocks the generous mark: absence of evidence becomes an explicit “no”, not a “partly”.
- “no overall mark”
- An average erases exactly the criterion you cared about, and it is the only figure anyone remembers.
What you get out of it
A reusable grid, and a judgement that can be contested point by point. The order is what counts: criteria written after reading describe what is already in front of you, and stop judging anything. It is the pocket version of what a serious evaluation does.
Going further
Keep the grid and run it over work from six months ago. If the marks come out the same everywhere, your criteria do not discriminate: they describe your taste, not quality.