Worksheet and differentiation prompts
Three-tier worksheets fail when the only difference is question count. Here is how to tier by cognitive level and support instead, with templates for language support and SEN adaptation.
以繁體中文閱讀On Wednesday period 1, a Primary 5 English teacher distributes a three-tier worksheet on the past tense that AI produced. The base tier has twelve gap-fill items, the core tier has ten gap-fills with harder verbs, and the extension tier has six gap-fills with adverbs. Mid-lesson she overhears a student mutter: "Extension is just more blanks." Eight students who need support are still stuck on item 3 of part one; the five who finished early are also just filling more gaps in the extension.
Her prompt had been one line: "Make a tiered past tense worksheet, basic, core and challenge." The problem is not the model. The prompt never separated cognitive level from amount of support. This is lesson 2 of module two, following lesson 6, lesson planning prompts: once the plan is set, the same lesson needs three entry points students can work through in parallel. For the instructional design background, see our column article differentiated instruction with AI. This lesson covers the prompts only.
What fake tiering looks like
The commonest failure: the three tiers differ only in question count or number size, with the same cognitive demand throughout. Students see through it quickly. Those who need support stay stuck on the same error type; those who finish early learn nothing new.
Real tiering pulls apart on two axes:
- Cognitive level: tier one stops at understanding and direct application, tier two applies in context, tier three analyses, compares, explains or creates.
- Amount of support: tier one supplies a step framework, a worked example and a word bank; tier two supplies part of the framework; tier three supplies none and the student chooses the method.
Same staircase — some people need the handrail, some need a landing halfway, some run straight up. You are not building three staircases of different heights.
Differentiation is not "the same item three times with uglier numbers". It is "the same learning focuses at three different depths of thinking".
Template 1: the three-tier worksheet
The example below uses Primary 5 English past tense; swap in your own unit:
You are a materials design assistant for a Hong Kong primary English Language teacher.
Task: design a three-tier worksheet.
Context: Primary 5, unit "past tense in narratives". Learning focus: (1) distinguish regular and irregular past forms; (2) use past tense in a short paragraph to sequence events; (3) check subject–verb agreement. Prior knowledge: present-tense narration. Class profile: 30 students, about 8 still confusing past forms, about 5 who finish the core quickly. Lesson 35 minutes; worksheet for the final 15 minutes plus homework.
Output format: three sections named part one, part two and extension — not high, middle or low. Each section gives a one-sentence cognitive level, learning focus numbers addressed, questions and support provided. End with an answer key and marking points per question.
Tier specification: part one — understanding and direct application, with a "time word → verb form" reference table, 4 questions. Part two — apply in an 80-word situational paragraph, first sentence supplied as a frame only, 5 questions. Extension — 3 questions, at least one asking the student to shift tense in a paragraph and explain one change, at least one asking the student to write a three-sentence past-tense story of their own.
Constraints: all three sections address the same learning focuses; extension difficulty must not come from longer words alone; mark "teacher to verify" if unsure whether a word is above level.
Thirty-second check: does the extension use explain, compare or write your own? If it is only more blanks, the tiering is fake.
Template 2: language support for non-Chinese speaking students
You are a materials adaptation assistant for a Hong Kong primary [subject] teacher.
Task: produce a language-support version of the worksheet below for non-Chinese speaking students to use in parallel.
Context: Primary 5, unit "[name]". Two non-Chinese speaking students, four years in the school, fluent orally, written vocabulary at about 60% of year-group level, struggling most with four-character expressions and formal written patterns.
Output format: question order exactly matching the original; a word bank before each question (hard word, everyday spoken equivalent, example sentence); a sentence frame with gaps for each writing question; a teacher notes column stating what was adapted in each question.
Constraints: do not lower the learning focus, only the language barrier; do not drop questions; example sentences from the daily life of Hong Kong primary students; mark "teacher to verify" if unsure whether a word is common at this level.
Original:
[paste]
Template 3: SEN adaptation
You are a special educational needs support assistant for a Hong Kong primary General Studies teacher.
Task: produce an adapted version of the inquiry worksheet below.
Context: Primary 5, unit "changes in materials". Three students with attention difficulties and one with dyslexia. Original has a 120-word experiment description and five questions.
Output format: (a) experiment description as numbered steps, one instruction per step, tick box before each; (b) same five questions with command verbs in bold and writing space between items; (c) a circling or drawing response option for the student with dyslexia; (d) teacher note explaining which need each adaptation serves.
Constraints: do not reduce content, only change presentation and response mode; each step instruction under 12 words; mark "teacher to verify" for any uncertain safety matter.
Original:
[paste]
Template 4: extension tier and self-assessment
For the five or six students who finish the core with time to spare:
You are a materials design assistant for a Hong Kong [primary / secondary] [subject] teacher.
Task: design an extension activity and a student self-assessment sheet.
Context: [year level], unit "[name]", learning focus: [list]. About six students finish the core with time to spare.
Output format: three extension tasks, each requiring analysis, comparison or creation, with success criteria; a five-row self-assessment sheet with one observable behaviour per row (for example "I can say why I chose this tense"), for students to tick after finishing.
Constraints: extension must not be more items of the same type; must be startable without further teacher explanation; self-assessment must not use unverifiable wording like "I understand now".
Where people go wrong
- Tiering by quantity. Twelve items versus six is workload. Name the cognitive level of each tier in the prompt. Self-check: are all three tiers the same item type? If yes, change the verb demand, not the count.
- Each tier on a different learning focus. That produces three homeworks that cannot run together. Self-check: are the learning focus numbers identical across tiers?
- Adaptation afterwards. Question order and wording drift; non-Chinese speaking students cannot follow the projection. Self-check: can adapted question numbers line up item by item with the original?
- Tier names that carry judgement. "Strong group / weak group" printed on the sheet is read by students. Self-check: does the output format specify neutral names?
Try it once
Time: 10 to 15 minutes.
What to do: Take your most recent untiered worksheet (or consolidation questions just generated in lesson 6). Write the template 1 prompt and generate three tiers. Check whether the extension includes at least one item requiring explain or design. If not, revise only the tier specification line in the prompt and regenerate the extension section.
What to submit: The three extension items and the tier specification line you changed. In one sentence, say how the extension differs from part one in thinking demand.
How you know it worked: You can explain to a colleague why fast finishers do the extension while students who need support stay on part one, and both are still assessed on the same learning focuses — meaning tiering opened cognitive depth and support, not a question-count trick.
With worksheets done, the next time sink is setting papers and writing rubrics: lesson 8, question setting and rubric prompts.
What to take from this lesson
- Name the cognitive level and amount of support for each tier in the prompt, instead of faking differentiation with question count alone.
- Require all three tiers to map to the same set of learning focuses so the lesson can run in parallel and class mastery can be compared afterwards.
- Produce the adapted version in the same prompt or the same workflow pass, so question order and wording stay aligned with the original.
- Design an extension tier that asks students to explain, compare or create, rather than using bigger numbers or longer calculation.
Questions that keep coming up
It depends on presentation. Many schools use neutral names such as part one, part two and extension, and let students move up after finishing the base section. Specify the naming in the prompt and avoid high, middle and low.
Check whether the extension tier uses verbs like explain, compare or design, and whether all three tiers map to the same learning focus numbers. Miss either and you usually have only quantity or number size differing.
Word banks, sentence frames, simplified instructions and visual suggestions are solid as a draft. Only the teacher knows actual language level, so the adapted version must be corrected for your class, especially vocabulary choice.
You can, but question order and wording often drift from the original and the lesson cannot run in parallel. Safer: request both in one prompt, or paste the full original into the adaptation prompt.