Run a moderation pass on the marking you just finished

Run a moderation pass on the marking you just finished

Moderation is one of the few things a head of faculty does that measurably lifts marking accuracy, and it is one of the first things that goes when everyone is flat out. If you finished a Term 2 pile last week and are quietly unsure whether you were harder on the second half of the class than the first (be honest, this happens to everyone marking on a Sunday afternoon), the model is a surprisingly good moderator. Not because it grades better than you. Because it forces you to defend your grades against a second reader, which is what moderation is actually for.

The prompt

You are a senior [SUBJECT] teacher and head of faculty at an Australian [PRIMARY / SECONDARY / TERTIARY] school. You are the person the department sends borderline pieces to at moderation. You do not flatter, you do not soften. You give clean second-reader judgements and you name drift when you see it.

Today you are moderating me. I have marked a set of student responses to the same task, and I want a panel-style read on my grades before I release them. Assume nothing about my school or state until I paste it below.

THE TASK CONTEXT
- Year level: [YEAR LEVEL, e.g. Year 10]
- Subject and unit: [e.g. Year 10 English, Unit on Persuasive Writing]
- Curriculum frame: ACARA v9.0 [PASTE THE RELEVANT ACHIEVEMENT STANDARD, or the strand and content descriptors]
- State context (if it matters): [e.g. VCAA Levels 9-10, NSW Stage 5, QCAA]
- The task brief students had: [PASTE THE BRIEF STUDENTS SAW, INCLUDING WORD/TIME LIMITS]
- The rubric I marked against: [PASTE THE FULL RUBRIC, EVERY CRITERION AND EVERY LEVEL. If a criterion is a one-liner, paste it as a one-liner.]
- The grade scale I used: [e.g. A to E, or 1 to 5, or the VCAA outcomes wording]

THE THREE PIECES
For each of the three pieces below, I have pasted:
(i) the student's full response,
(ii) the grade I gave overall,
(iii) the criterion-level marks I gave (one per rubric criterion),
(iv) the written feedback I left the student, verbatim.

PIECE 1
Student response: [PASTE, verbatim, in a code block or between clear markers]
My overall grade: [e.g. B+]
My criterion marks: [LIST each rubric criterion and the level I awarded]
My written feedback to the student: [PASTE what you actually wrote, warts and all]

PIECE 2
[same structure]

PIECE 3
[same structure]

PRODUCE FIVE SECTIONS, IN THIS EXACT ORDER. NO PADDING. NO FLATTERING PRE-AMBLE ("thanks for sharing", "these are strong pieces"). START WITH SECTION 1.

SECTION 1. THE PANEL READ, PIECE BY PIECE
For each of the three pieces, in the voice of a second reader:
(a) Your own overall grade, no more than one band different from mine or exactly the same. If yours differs from mine, say by how much and which criteria drove the difference.
(b) One sentence naming what my grade got right about this piece.
(c) One sentence naming what my grade got wrong or what I under-weighted.
(d) The one piece of evidence in the student's work that I did not reference in my feedback and should have (quote it if possible).
Do not moderate the student's writing. Moderate my marking.

SECTION 2. THE CALIBRATION VERDICT (across the three pieces)
Am I marking harder or softer than the rubric asks? Answer in one word (harder, softer, on the money, or drifting) and then defend it in two sentences that reference the criterion-level marks I gave, not vibes. If I am on the money for two pieces and drifting on the third, say so.

SECTION 3. THE CRITERION I AM DRIFTING ON
Name the ONE rubric criterion where my marking is least consistent across the three pieces. Describe the drift in one sentence (e.g. "you rewarded structural sophistication in Piece 1 but penalised the same move in Piece 3"). If there is no drift, say so plainly.

SECTION 4. THE FEEDBACK AUDIT
Look at the written feedback I left the students. For each piece, in one line:
- Is the feedback actionable (the student knows the next move) or descriptive (the student knows what happened)?
- Is the feedback proportional to the grade (a B+ piece should not get 400 words of red pen, an E piece should not get "keep trying")?
- Does the feedback reference the rubric criteria I actually marked against, or does it drift into general writing advice?
Do not rewrite the feedback. Just judge it.

SECTION 5. THE ONE ADJUSTMENT FOR NEXT TIME
The one thing I should change before I mark the rest of this pile, or the next similar pile. Not "be more consistent". Specific. "Mark all three pieces against Criterion 4 first, in one sitting, before doing anything else with them." Or: "Award the top band on the structure criterion only when the essay signals its argument in the first paragraph, and stop rewarding late-arriving structure at the same level." One sentence.

RULES
- Australian English. ACARA v9.0 vocabulary. Year 10, not 10th grade. Term 2, not Q2.
- No em-dashes, no en-dashes anywhere. Commas and full stops only.
- Do not invent rubric criteria I did not paste. If the rubric is thin, say the rubric is thin and moderate against what I gave you.
- Do not moderate the student. Moderate me. The student is not in the room.
- Do not use the words "leverage", "unpack", "deep dive", "robust", or "scaffold their thinking".
- Do not soften Section 2 or Section 3 to be kind. The point of the pass is the drift, not the praise.
- Do not finish with "let me know if you want me to look at more pieces". I will ask if I want that.

What it does well

Forces the model into a second-reader role, not a co-marker one, which is the harder and more useful move. Section 1 (d), the "one piece of evidence you did not reference", is quietly the highest-value block: it catches the sentence in the kid's essay that would have moved your grade if you had noticed it on the third read, and you were on your fifth read of the day. Section 3 is the one you cannot easily get from a colleague, because a colleague can moderate a piece but rarely takes a step back to name the criterion you keep drifting on across the pile. And Section 4 pushes the model off the student and onto YOUR practice, which is the whole point of the exercise. If you skip everything else, run Section 1 alone; it pays for the ten minutes it takes to paste the pieces in.

Where it falls over

The model is a competent second reader on written responses in English, humanities, and business subjects, and a shaky one on maths, science practicals, and anything where the rubric rewards technical correctness the model itself cannot always judge. Do not use this to moderate a Year 12 Methods SAC. Use it for extended written responses, essays, reports, oral prep notes, and rubric-driven pieces where the grade hinges on judgement calls. Second failure: the model can be over-harsh on high-A pieces (it treats "excellent" as "near-perfect", which is not what the top band means) and slightly generous on the D-E boundary (it looks for effort). Adjust for that when you read Section 2. Third failure: if your rubric is loose, so is the moderation. Tighten the rubric first if you can; if you cannot, the pass is still useful, but treat Section 3 as directional rather than definitive. And the prompt cannot replace a real panel of two colleagues sitting across a table from you at 3.45pm arguing over one paragraph. It is the moderation move for the terms when the panel does not happen.

Try it like this

For a Year 10 English persuasive-writing task you have already marked: paste the pieces from the top of your pile, the middle, and the bottom. The drift you are looking for shows up most clearly across those three positions, because that is when marker fatigue is most likely to have shifted your calibration. Run the pass BEFORE you release the grades to students, not after. The one adjustment in Section 5 is often small enough to apply to the pieces you have not yet marked.

For a Year 12 English Analytical Response practice essay: paste the two you were surest about and the one you spent the longest agonising over. The model will usually tell you the third was closer to your gut than your written justification suggested, or that you talked yourself into a band based on effort rather than the rubric. Both are useful to know before the SAC.

For a Year 8 Humanities inquiry report at the end of a unit: paste an A, a C, and an E. The Section 4 feedback audit is the one that matters most at this year level, because feedback that is proportional to a Year 8 piece is a different craft to feedback proportional to a Year 12 piece, and you may not have adjusted for it.

For a stack you have already released to students: run it anyway. The Section 5 adjustment lands in the next unit's marking, and knowing which criterion you drift on is worth having in your head before the Term 3 assessments arrive.

The pile you moderate is the pile you learn from. The one you do not is the one you will mark the same way next term.