Standard-setting comparison

Compare your work against examples at several grades, so you can see what the difference between bands actually is.

When to use this: When checking yourself
Learning impact: Reliable gain
Learning evidence: Credible with caveats
Where you are with the topic: Checking readiness
How often to use it: Worth repeating often

When to use it

Rubric language like “shows evaluation” still feels abstract.
You think you are mid-band and markers keep saying you are missing one move.
You have never seen what a bare pass, mid-band, and top-band version of the same brief look like side by side.

What to ask AI

Paste your work and the criteria. Ask for three versions of a comparable piece: bare pass, mid-band, top, without changing your content, show the same material at each level.

Ask which yours most resembles and why.

Ask for the specific difference between mid and top, sentences that exist in one and not the other.

Or ask for three unlabelled versions, rank them yourself, then check whether you ranked correctly.

How it helps

Rubrics describe bands in abstract language. “Demonstrates critical evaluation” and “shows some evaluation” are hard to tell apart until you have seen both.

Exemplars make the difference visible because the gap between bands is usually concrete and small, not more content, but a specific move.

In essays, the top-band version often says why the evidence it just gave is not conclusive, and the mid-band version does not.

Do this before rubric self-assessment if you have never seen exemplars: the rubric means little without them.

How to run it

Put your work and the marking criteria in front of you.
Get three comparable versions at different bands (prefer real institutional exemplars when available).
Rank unlabelled versions yourself before you are told the order.
Ask for the specific mid-vs-top difference in sentences, not slogans like “more analysis.”
Name the one move your piece is missing, then practise producing that move.
If you can see the difference but cannot produce it, switch to draft-and-critique loops.

Example exchange

You

I've got an essay paragraph and the marking criteria. Can you show me three comparable paragraphs at different bands, unlabelled, so I can rank them?

Here are A, B, and C. Rank them and say what you notice. I won't tip the intended order yet.

You

B, then A, then C. B admits a limit on the evidence; A just asserts; C only describes.

That's the right order. Yours looks most like A. The top-band sentence in B is the one that weakens its own case on purpose, can you try that move in yours?

Copyable prompt

Here's my work and the marking criteria. Produce three versions of a comparable
piece: one at a bare pass, one mid-band, one at the top of the scale. Don't
change my content: show me what the same material looks like at each level.
Then tell me which mine most resembles and why.
Between the mid-band and the top-band version, what is the SPECIFIC difference?
Not "more analysis": what sentences exist in one and not the other, and what do
they do?

The Tell

You can see the difference between bands and still cannot produce the top-band move.

That is a skill gap, not a knowledge-of-criteria gap, switch to draft-and-critique practice.

Treating generated exemplars as official standards also fails: use them for shape, real exemplars for authority.

Principle evidence

Strength of the underlying learning idea, not a claim about AI products.

The underlying learning idea is rated moderate. Comparing work to grade exemplars is standard-setting and assessment literacy (worked examples of quality levels). Seeing criteria instantiated is the load-bearing move.

AI delivery evidence

Whether an AI tutor delivers this method well is a separate question.

AI-chosen exemplars and grade labels are speculative. Prefer real institutional exemplars when available; treat generated ones as guides to the shape of the difference.