Differentiate one task into three tiers without doubling your prep

Differentiate one task into three tiers without doubling your prep

Differentiation is the piece of good practice every teacher agrees with and nobody has time to actually do. The Sunday-night rewrite of Monday's worksheet into three versions is where the plan quietly falls over, and the failure mode is not the intent, it is the forty minutes. The version of this everyone tries with AI ("make this easier for my strugglers") comes back as a completely different task on a completely different topic, which now does not match the learning intention, which now cannot be marked against the same success criteria. This prompt is the constrained version. Same LI. Same success criteria. Same stimulus. Same lesson time. Three access points, all defensible in front of a coach or a parent, none obviously flagged to the student sitting next to their neighbour on the higher tier.

The prompt

You are helping me differentiate one existing task into three tiered versions for the same [YEAR LEVEL] [SUBJECT] class in an Australian school. The three versions must all target the SAME learning intention and produce evidence I can mark against the SAME success criteria. I do not want three different tasks. I want one task, three access points.

You are working in the tradition of Tomlinson on tiered instruction (same big idea, different route), Willingham and Sweller on cognitive load (more scaffolding, not easier content), and the ACARA v9.0 principle that differentiation adjusts the pathway to a standard, not the standard itself.

CONTEXT I WILL PROVIDE

- Year level and subject: [E.G. Year 7 Maths, or VCE Unit 3 Biology]
- The learning intention (LI) for this lesson: [WHAT THEY SHOULD BE ABLE TO DO BY THE END OF THE LESSON, IN ONE SENTENCE]
- The success criteria I will mark against: [2 TO 4 CRITERIA, IN STUDENT-FACING LANGUAGE]
- The original task, exactly as it currently exists: [PASTE THE TASK, WORKSHEET, QUESTION, PROMPT OR ACTIVITY. INCLUDE THE STIMULUS AND ANY SCAFFOLDS ALREADY IN IT.]
- Anything I already know about my cohort: [OPTIONAL, E.G. "6 EAL LEARNERS, 2 ON A READING INTERVENTION, 3 WORKING ABOVE STANDARD" OR "MOST OF THE CLASS HAS SEEN QUADRATICS BUT ONLY 4 CAN FACTORISE FLUENTLY"]

WHAT I NEED YOU TO PRODUCE

Three tiered versions of the SAME task. All three must:
- Produce work I can mark against the SAME success criteria
- Take roughly the SAME time to complete in a lesson
- Use the SAME stimulus, context and content (do not swap the topic to something easier or fancier)
- Look near-identical on the page (I do not want three visibly different worksheets that mark students out to each other in the room)

The three tiers are:

1. ACCESS
The same task, with more scaffolding for a student whose prerequisite skills are patchy or whose cognitive load needs reducing. Add scaffolds by:
- Chunking multi-step questions into labelled steps
- Providing one worked example at the top of the task, faded (all steps shown for example 1, one step blanked out for example 2, two steps blanked out for example 3)
- Adding a sentence starter, a key-vocab strip, or a formula or definitions box at the point of need
- Reducing extraneous stimulus load (unnecessary decorative text, ambiguous phrasing), not the standard
Do NOT lower the standard. Do NOT swap the content for an easier topic. The access version still evidences the LI.

2. ON-STANDARD
The task as it should sit for a student working at year-level standard. Take my original task and tighten it (cut unclear wording, fix any ambiguous instruction, make the success criteria visible on the task itself). If my original is already clean, return it near-unchanged and say so in a one-line note.

3. STRETCH
The same task, extended for a student who has the LI already and needs to go deeper. Extend by:
- Adding a "now justify", "now compare", or "now generalise" step that requires the student to explain WHY their answer works, or to apply it to a second case
- Introducing one non-standard variation of the stimulus (a boundary case, an ambiguous input, a real-world constraint)
- Requiring the student to critique or improve a given answer, not only produce one
Do NOT just add more of the same questions. Extension is not volume. It is depth on the same task.

Return the three versions as three clearly labelled blocks, in this order: ACCESS, ON-STANDARD, STRETCH. Format each as a print-ready task, not a description of a task. I should be able to paste each block straight into a Google Doc, adjust the header, and print.

Then, in a SEPARATE block below the three tasks, and clearly labelled "FOR MY EYES ONLY, DO NOT PRINT":

- One sentence naming the SUCCESS CRITERIA row that the access version might not fully evidence, and how to check for it in the student's work
- One sentence naming what a student on the stretch version will produce that the on-standard version does not, so I know what "stretch-worthy" work looks like when I mark it
- One sentence flagging any student in the cohort context I gave you who might sit AWKWARDLY between two tiers, which tier to hand them and why
- One sentence on the piece of the original task I should REWRITE for next term, based on whatever was unclear enough to need tightening in the on-standard version

RULES

- Australian English throughout. Realise, colour, behaviour, analyse. Year 7, not 7th grade.
- No em-dashes and no en-dashes anywhere. Commas, full stops, parentheses, semicolons only. Regular hyphens in compound words are fine.
- No "differentiated", "scaffolded", "extension" or "challenge" labels on the student-facing task. Students should not be able to tell which tier they are on. Label the three blocks for ME at the top, but the printable task itself does not name the tier.
- No praise stickers, no "great job!" language baked into the task. This is a task, not a certificate.
- If the LI or success criteria I gave you are too vague to differentiate against (e.g. "understand fractions", "engage with the text"), STOP and ask me to sharpen them before drafting. A vague LI produces three near-identical tasks and wastes both our time.
- If the original task I pasted does not actually assess the LI I gave you (a common trap), STOP and tell me. Fixing the task-to-LI match matters more than tiering the wrong task.

If any of the following are true, STOP and tell me before drafting:
- The task is a summative assessment (SAC, NAPLAN, VCAA-style exam, standardised test). Tiering a summative changes the standard, which is a different job that belongs with the head of learning area, not with me and you on a Sunday.
- The task is a group activity as pasted. Tiering group activities is a different design problem; tell me and I will reshape it as individual work first, or you can suggest role differentiation within the group instead.
- The learning intention names a specific ACARA content descriptor, elaboration or VCE study-design outcome and the on-standard task does not actually meet it. In that case, fix the on-standard version to meet the descriptor first, then tier.

What it does well

The forcing constraints do the work. Locking the LI, the success criteria and the stimulus stops the model doing what teachers actually hate about "make this easier", which is quietly swapping out the topic for something easier to talk about. Requiring the three versions to look near-identical on the page is what protects the kid on the access tier from being visibly marked out to their neighbour, and the faded-worked-example scaffold in tier one is straight Rosenshine, so the access version teaches, not just simplifies. The "for my eyes only" block at the end is the sleeper feature: it tells you which success-criteria row the access version might not fully evidence, so you know where to sit next to that kid when they hand it in.

Where it falls over

Two honest limits. First, the model has never met your students. Its guess at who sits awkwardly between tiers is only as sharp as the cohort context you paste, so if you type "mixed ability", you will get mixed-ability advice back. Second, on some topics (especially early-primary numeracy, and VCE-specific study-design content) it will over-scaffold the access version to the point where it does the thinking for the student, not for the student. Read the access version against your success criteria before printing, and if it evidences the LI without the student actually doing the work, tighten step 3 or step 4 by hand.

Try it like this

For Year 7 Maths, adding and subtracting fractions with unlike denominators: paste the four-question worksheet you already use, the LI ("I can add and subtract fractions with unlike denominators by finding a common denominator"), and the three success criteria. You will get an access version with a faded worked example on top, an on-standard version that tightens the wording of question 3 (there is always a question 3), and a stretch version that asks the student to explain why the common denominator method works and to invent a fraction pair with no small common denominator.

For Year 10 English, a persuasive writing response to a stimulus editorial: paste the original prompt, the LI ("I can construct a persuasive paragraph using a claim, evidence from the text, and analysis"), and the rubric rows. The access version arrives with a sentence-starter strip and a two-column planning grid. The stretch version keeps the same stimulus but asks the student to write TWO paragraphs from opposing positions, then a one-sentence note on which was harder to argue and why.

For VCE Unit 3 Chemistry, an equilibrium calculation question: paste the item and the study-design outcome. The access version breaks the calculation into labelled steps with the ICE table set up for them. The stretch version keeps the same numbers but adds a "if temperature increases by 10 degrees, predict and justify" follow-up, which pushes into Le Chatelier without changing the task.

If you only have five minutes: skip the cohort context, paste the task and the LI only, and ask for the access tier alone. That is the one that saves the most planning time and the most reading of confused faces on Monday.

Print the three, keep the "for my eyes only" block on your desk, and mark all three stacks against the same success criteria on Tuesday morning.