Differentiate a reading across three levels without three preps

Differentiate a reading across three levels without three preps

Reading differentiation is the thing every Australian teacher of a mixed-ability class is quietly doing badly, because doing it properly means writing three parallel versions of every text and nobody has that Wednesday night in them. The consequence is the reading that sits above five kids and below ten, which teaches almost no one. Second week of the winter holidays is a good time to build the version of this prompt you will use for the rest of Term 3, on the readings you know you are going to set anyway.

The prompt

You are a senior [SUBJECT] teacher in an Australian [PRIMARY / SECONDARY / TERTIARY] school. You have taught this year level for at least a decade and you have taught it to a full mixed-ability spread including EAL/D students and students with reading disabilities. You differentiate reading by rewriting the text, not by giving weaker readers less of it.

Today you are going to produce three parallel versions of one source text I paste below, so I can use the same content with a mixed-ability class in the same lesson.

THE CONTEXT
- Year level of the class: [YEAR LEVEL, e.g. Year 8]
- Subject and topic: [e.g. Year 8 Humanities, Rome as a case study of ancient governance]
- Curriculum frame: ACARA v9.0 [PASTE THE RELEVANT CONTENT DESCRIPTOR OR STRAND]
- State context (if it matters): [e.g. VCAA Levels 7-8, NSW Stage 4, QCAA]
- Class composition in one line: [e.g. "26 students, four EAL/D at intermediate level, three with reading adjustments (~2 years below), two identified as extension"]
- Purpose the students will use the reading for in the same lesson: [e.g. "answer six comprehension questions and start a paragraph response", or "extract three pieces of evidence for a debate", or "annotate for author bias"]

THE SOURCE TEXT
[PASTE THE FULL SOURCE TEXT, verbatim, between clear markers or in a code block. Do NOT summarise it before pasting. If it is more than 800 words, paste the section you actually want the class to read, not the whole article.]

PRODUCE THE FOLLOWING SECTIONS, IN THIS EXACT ORDER. NO PADDING. NO FLATTERING PRE-AMBLE ("great source", "rich text"). START WITH SECTION 1.

SECTION 1. THE SCAFFOLD VERSION (target reading age ~2 years below the class)
A rewritten version of the source that keeps every key idea, every argument move, and every piece of evidence the mainstream version has. What you change: sentence length (shorter, one idea per sentence), sentence connectives (make cause and contrast signals explicit: "so", "but", "even though"), pronoun distance (name the referent again when a pronoun would sit more than one clause from its antecedent), and vocabulary at the concept-neutral level. What you do NOT change: any subject-specific technical term (patrician, plebeian, aqueduct, judicial review, mitosis, allegory). Keep those. Gloss them in parentheses on first use, once. Length: roughly 70% of the mainstream version.

SECTION 2. THE MAINSTREAM VERSION (target reading age at the class year level)
A rewritten version of the source pitched at the median reader in the class. Cut the source down to the load-bearing bits (drop the newspaper hook, drop the throwaway anecdote at the end) so the class can read it inside 8 to 10 minutes at year-level pace. Keep sentence variety. Keep at least one long sentence with subordinate clauses per paragraph, because that is the sentence structure the mainstream reader needs practice with, not to have removed for them. Length: whatever the source needed cut to; roughly 400 to 550 words for a standard 60-minute lesson.

SECTION 3. THE STRETCH VERSION (target reading age ~2 years above the class)
A rewritten version of the source that ADDS to it, does not simplify further. What you add: one embedded counter-position or complicating source (real if you can name it, invented and labelled "hypothetical counter-source" if you cannot), one additional layer of vocabulary that the extension reader can handle (e.g. "ratify", "populist", "oligarchic", "photolysis", "unreliable narrator"), and one longer sentence with two subordinate clauses that the mainstream version simplified. Length: roughly 130% of the mainstream version. Do NOT add new content the class will not otherwise cover; add difficulty inside the same territory.

SECTION 4. THE SHARED COMPREHENSION QUESTIONS
Six questions that a student at any of the three levels can answer, and that produce answers that let me tell whether they engaged with the actual content, not just whether they can decode. Order: two literal-comprehension questions, two inference questions, two evaluation or judgement questions. Every question must be answerable from ALL THREE versions. If a question can only be answered from the stretch version, rewrite it or move the content down into all three versions. Below the six questions, list the ONE extension question that is only answerable from the stretch version, labelled clearly.

SECTION 5. THE VOCABULARY BRIDGE
List the five to seven words in the mainstream version that an EAL/D or reading-adjustment student needs supported to read it without the scaffold version. For each word: (a) the word, (b) the one-sentence plain-English meaning in the sense the source uses it, (c) the sentence from the mainstream version it appears in. This is for the teacher to pre-teach on the board in three minutes at the start of the lesson, so the scaffold version is not the only support option.

SECTION 6. THE HONEST NOTES
Two to four short bullets I need to see before I use these versions.
- Any point in the source where the scaffold version has quietly lost meaning that a purist would object to (e.g. "the scaffold version collapses the distinction between causation and correlation in paragraph 3; if the lesson objective is about causation, use the mainstream version").
- Any assumption you made because the source was ambiguous, and what you would need to know to tighten it.
- Any student the composition line suggests will still not access the scaffold version, and what a fourth support (audio, paired reading, comprehension prompts read aloud first) would look like.

RULES
- Australian English throughout. Use Australian examples where the source is a general one and you have to invent an example ("Melbourne suburb", not "Chicago neighbourhood"). Never Americanise a source that is not Australian; keep its original context and just rewrite the language.
- No em-dashes and no en-dashes anywhere. Commas and full stops only.
- Do not add content to any version beyond what the source contains, except in Section 3 where the addition is explicit and labelled.
- Do not rewrite the source in a way that softens its argument or removes its politics. If the source is about colonisation, the scaffold version is still about colonisation; you do not turn it into "settlement".
- Do not use the words "leverage", "unpack", "deep dive", "robust", or "scaffold their thinking".
- Do not finish with "let me know if you want more levels". I will ask if I want more.

What it does well

Forces the model to preserve subject-specific vocabulary across all three versions, which is the thing that most classroom differentiation quietly gets wrong. The scaffold version of a Year 8 Humanities reading that removes "patrician" and "plebeian" has removed the point of the reading; the good version keeps the words and props them up. Section 4 is the block that pays for the whole thing: shared questions across three parallel versions is what makes the reading feel like the same lesson to students, not three ability groups sitting in the same room. And Section 6 is the honest audit almost no teaching-strategy document includes: where the differentiation itself has cost something, and what the fourth support looks like for the student the three versions still will not reach.

Where it falls over

The model can produce a stretch version that is longer and harder without being genuinely richer. Read Section 3 with a red pen; if the added "counter-position" is just a paragraph of "however, some historians argue" with no actual second source or second lens, ask the model to name the source it is drawing on or to remove the addition. The second failure is in Section 5: the vocabulary bridge is often too short. If your class has five EAL/D students and you can only see four words listed, run the mainstream version past the model again and ask "which words has an intermediate EAL/D reader not seen this year". Third failure: the model will sometimes give you three versions that are the same information at three lengths, not three genuinely different reading experiences. If Section 1 and Section 3 read like the same voice, ask specifically for "shorter connective sentences and named referents in Section 1" and "one subordinated sentence per paragraph in Section 3". And the prompt cannot judge whether the source is worth differentiating in the first place. If the reading was already thin, three thin versions is not progress.

Try it like this

For a Year 8 Humanities reading on ancient Rome: paste a chapter extract from your textbook, set the composition to include EAL/D and reading-adjustment students, and let the model produce all six sections. Print all three versions on the same paper stock, in the same font, with the same heading, so students cannot tell from across the room which they were handed. That is the delivery move that makes differentiation stick.

For a Year 10 English text-study lesson working from a critical essay on the novel: run the prompt but tell the model to keep the essay's argument intact across all three versions. The scaffold version keeps the thesis; it just makes the paragraph structure easier to follow. Do not use it to change the argument.

For a Year 9 Science reading on cellular respiration: keep every technical term (glycolysis, mitochondria, ATP) in all three versions and let the model gloss them once in the scaffold version. The vocabulary bridge in Section 5 becomes the board work for the first three minutes of the lesson, which is exactly the right length for pre-teaching before the class reads.

For a primary Year 5 History reading on the First Fleet from a state DET resource: set the composition line to name the EAL/D students explicitly and specify that Section 6 should suggest whether audio narration is the fourth support you need for one student. Section 4 comprehension questions will be your literacy block for the day.

If you only have ten minutes on Sunday night: run Section 1 and Section 4 only, and use the source you already have as the mainstream version. That is enough to get you through Monday without the reading-level mismatch that produces the "I'm done, miss" moment before the rest of the class has finished paragraph two.

The reading you write once at three levels is the reading you re-use for a term. The one you write at one level is the one three of your students never really read.