The English Tutor Handbook · Volume 0013 · Series ID SETC-ETH-0013
A worksheet is never just a worksheet.
It is a measurement instrument.
It is a practice environment.
It is a set of hidden assumptions about what the learner already knows, what the learner must notice, what the learner has to remember, and which decisions the learner must make before an answer can appear.
If the material does not isolate the capability clearly enough, the tutor may end up teaching the wrong thing with great confidence.
This volume is about English task and material design. It is not a generic resource-selection article and it is not a new curriculum owner. Its job is practitioner-side: how an English tutor chooses, edits or constructs texts, questions, examples and practice so that the learner’s performance can be interpreted and the next teaching decision becomes clearer.
Quick Read
- Start from the capability, not from the worksheet.
- Know which variables the material changes.
- Separate target difficulty from accidental difficulty.
- Use short tasks for diagnosis and high attempt density.
- Use longer authentic tasks for integration and transfer.
- Build near-miss options when judgement is the target.
- Change one meaningful variable at a time when diagnosing.
- Make materials difficult for the reason you intend.
- Retire materials that produce familiar success without new evidence.
1. Begin With the Decision, Not the Resource Folder
Tutors often plan by searching for a worksheet on a topic.
“Inference worksheet.”
“Composition practice.”
“Editing exercise.”
Those labels are too broad to guarantee useful practice.
Start with a decision specification instead:
The learner must compare two plausible interpretations and decide which one is best supported by the whole passage.
Now the tutor can ask whether the material actually creates that decision.
2. The Material Has a Hidden Task Model
Every English task makes demands in several layers:
- language access;
- background knowledge;
- memory;
- attention;
- reasoning;
- response construction;
- writing or speaking control;
- time management.
The tutor should know which demands are intentional and which are collateral.
If the target is pronoun reference, a passage with unusually difficult specialist vocabulary may muddy the evidence. If the target is timed synthesis, higher load may be exactly the point.
3. Target Difficulty and Accidental Difficulty
Target difficulty is the challenge the tutor wants the learner to confront.
Accidental difficulty is challenge added by the material that is unrelated to the target.
If Iona is practising evidence discrimination, the tutor may deliberately include two near-relevant evidence candidates. That is target difficulty.
If the same passage also contains five rare technical terms that block basic understanding, the evidence-discrimination result becomes hard to interpret. That may be accidental difficulty.
4. The Clean Diagnostic Task
Diagnosis benefits from cleaner tasks than performance simulation.
A clean diagnostic task tries to minimise irrelevant variables so the tutor can inspect one mechanism.
Examples:
- a short paragraph to test reference tracking;
- two nearly identical answers to test completeness judgement;
- a single paragraph plan to test architecture;
- three near-synonyms to test register;
- one listening segment with changed information to test update control.
Once the mechanism is understood, the tutor can restore complexity.
5. The Authentic Integration Task
Clean tasks are not enough.
Eventually the learner must use the capability inside a whole reading passage, composition, oral response, listening paper or mixed examination task.
Authentic tasks test integration:
- Can the learner recognise the need for the skill?
- Can the learner retrieve it without a label?
- Can the learner operate while other demands compete?
- Can the learner preserve it under realistic time?
The tutor should move between clean and authentic tasks rather than treating one as universally superior.
6. Build Backward From the Wrong Answer You Need to Distinguish
Good material can distinguish between plausible learner states.
If the tutor wants to know whether the learner understands cause versus sequence, include an answer that is chronologically related but not causal.
If the tutor wants to test register, include a synonym that is semantically accurate but socially inappropriate.
If the tutor wants to test completeness, include an answer containing one correct detail but missing the required relationship.
The distractor should represent a real misconception, not random wrongness.
7. Near-Miss Design
Near misses are particularly powerful once basic knowledge is secure.
They force discrimination.
- relevant versus sufficient evidence;
- grammatical versus clear sentence;
- precise versus overstrong vocabulary;
- developed versus repetitive paragraph;
- possible versus best-supported inference;
- formal versus appropriately formal tone.
A tutor who builds near misses is designing judgement, not merely correctness.
8. Keep One Variable Stable When You Need Interpretation
If two tasks differ in every way, a change in performance is difficult to explain.
When diagnosis matters, preserve most conditions and change one useful variable.
Same passage, different question scope.
Same vocabulary, different syntax.
Same paragraph content, different organisation.
Same listening content, changed note frame.
Controlled contrast creates better evidence.
9. Change the Surface When Transfer Matters
After the skill is understood, controlled similarity becomes a risk.
Change topic, genre, wording, evidence location, mode or time.
SETC-ETH-0008 owns the Unseen Transfer Test. Material design supports that test by providing surfaces different enough to remove familiarity while preserving the underlying decision.
10. Short Tasks Increase Attempt Density
A full essay may contain only one opportunity to practise paragraph-order decisions.
Six miniature planning tasks can create six decisions in the same time.
A full comprehension passage may contain one difficult reference chain.
Five short excerpts can create five reference decisions.
Short tasks are useful for repair, discrimination and dense practice.
They should later be reintegrated into full performance.
11. Long Tasks Test Coordination
Whole tasks reveal whether the learner can coordinate multiple processes.
A full writing task tests planning, development, sentence production, vocabulary retrieval, revision and time.
A full reading paper tests task interpretation, passage model, evidence selection, answer construction and pacing.
Use whole tasks when the question is integration, not when the tutor still needs to isolate one weak mechanism.
12. Text Selection for Reading
Select texts for the feature the learner needs to practise.
- reference chains;
- cause and consequence;
- contrast and qualification;
- multiple perspectives;
- implicit motive;
- paragraph structure;
- unfamiliar vocabulary that remains inferable;
- distributed evidence.
A famous or entertaining text is not automatically the right teaching text.
13. Question Design for Reading
The question controls what part of reading becomes visible.
A literal question tests retrieval differently from an inference question.
A “how do you know?” follow-up tests evidence control.
A “what would weaken your answer?” follow-up tests alternative handling.
Build question families that reveal different layers instead of treating all comprehension marks as one skill.
14. Material Design for Vocabulary
Vocabulary material should vary what the learner must do with the word.
- recognise meaning;
- infer meaning;
- choose a near synonym;
- reject a poor collocation;
- adjust register;
- retrieve from meaning;
- deploy in writing;
- use in speech.
A list that tests only definition recall gives limited evidence about usable vocabulary.
15. Material Design for Writing
Full compositions are necessary but expensive.
Build materials at multiple scales:
- idea generation;
- thesis comparison;
- paragraph ordering;
- evidence–explanation pairs;
- sentence combining;
- tone adjustment;
- revision choices;
- whole-piece integration.
The US Institute of Education Sciences’ writing guidance recommends explicit strategy instruction and Model–Practice–Reflect cycles. Materials should therefore let the learner practise the target strategy, not merely produce another complete composition.
16. Material Design for Editing
Editing tasks should distinguish three operations:
detect → classify → repair.
Marking the error location tests repair more than detection.
Giving the error category tests repair more than classification.
Leaving everything unmarked tests all three.
The tutor should know which task is being set.
17. Material Design for Listening
Listening material can vary:
- voice;
- speed;
- topic familiarity;
- information density;
- corrections and changed information;
- note-taking demands;
- distractor strength.
If the target is changed information, make that the main difficulty. Do not simultaneously introduce extreme speed and specialist vocabulary unless the point is higher-load transfer.
18. Material Design for Speaking
Speaking materials should generate thought, not invite memorisation.
- related but non-identical prompts;
- unexpected follow-ups;
- different audience conditions;
- limited preparation;
- required examples;
- counter-positions;
- short and extended response versions.
Repeated identical prompts can create fluent familiarity rather than flexible speaking.
19. Build Distractors From Learner Errors
The learner’s own mistakes are a powerful source of material design.
If several students choose evidence that is related but incomplete, build future contrast sets around related-versus-complete proof.
If students overuse formal vocabulary, build register discrimination sets.
If students confuse changed information in listening, build examples where first-state and final-state details compete.
Practice becomes responsive to real error signatures.
20. Build Materials With an Exit Route
If a scaffolded worksheet is useful today, decide what replaces it tomorrow.
Highlighted evidence → unhighlighted evidence.
Sentence starter → keyword cue → no cue.
Paragraph frame → purpose labels → blank page.
Materials should support withdrawal, not permanent dependence.
21. The Difficulty Dial
A material can be made harder in different ways.
- less explicit evidence;
- closer alternatives;
- longer dependency distance;
- more unfamiliar topic;
- higher language density;
- less support;
- more integrated skills;
- less time.
Turn one dial deliberately rather than pushing every dial at once.
22. The Material Should Produce Interpretable Failure
A task is educationally useful when failure tells the tutor something.
If the learner fails because the task contains too many uncontrolled demands, the failure may be dramatic but uninformative.
World-class material is not merely hard. It is well-specified.
23. The Material Should Produce Useful Success Too
A correct answer should reveal more than “the student got it right”.
Can the student explain the decision?
Can support be reduced?
Can the surface change?
Can the student reject a near miss?
Good material gives success somewhere to go.
24. Do Not Confuse Attractive With Instructionally Useful
Beautiful formatting can help attention.
It can also hide a poorly designed task.
Ask what the learner must think about.
If the answer is unclear, decoration is not the main problem.
25. Do Not Confuse Real-World With Automatically Better
Authentic texts can be rich, motivating and realistic.
They can also contain uncontrolled language and knowledge demands that make one mechanism hard to isolate.
Use authentic materials when authenticity serves the learning question.
Use adapted materials when control matters.
Return to authentic material before claiming transfer.
26. Maren: Material That Exposes Architecture
Maren does not need another page of sentence-level style practice.
Give her sets of strong sentences that can be arranged into multiple paragraph structures.
Ask which order produces progression and why.
27. Iona: Material That Exposes Evidence Competition
Choose passages where two evidence candidates are both plausible but differ in completeness or timing.
Do not rely on obviously wrong distractors.
Her target is judgement.
28. Leonie: Material That Exposes Performance Cost
Use paired tasks with similar language demand but different time conditions.
Observe what disappears when time compresses.
Then build practice around preserving the highest-value process.
29. Materials in a Three-Student Tutorial
One shared text can support three different learning jobs.
- Maren maps paragraph function.
- Iona compares evidence.
- Leonie works under a time gate and protects verification.
The material is shared. The question layer differs.
This preserves group coherence without pretending all learners need the same task resolution.
30. Research Anchors
The US Institute of Education Sciences provides recent toolkits for differentiated small-group reading instruction that emphasise data-driven decision rules, high-quality instructional materials aligned to learner needs, explicit instruction and purposeful independent activity. Its writing and reading practice guides likewise emphasise explicit strategy teaching, formative assessment and purposeful text use.
The Education Endowment Foundation’s work on worked examples, scaffolding and cognitive science highlights the value of carefully controlling task load, fading support and sequencing examples toward independence. Those principles support a material-design discipline in which tutors know what a task is loading and why.
31. How this guide fits
- Practice Engineering covers practice sequencing through retrieval, variation and mixing.
- The Unseen Transfer Test covers transfer verification.
- Exemplars Without Imitation covers model-answer and worked-example use.
This guide focuses on English task and material construction: choosing or building texts, questions and practice so target difficulty is deliberate, failure is interpretable and the task reveals the capability the tutor actually intends to teach.
32. The Task Laboratory Audit
- What exact capability is this task meant to reveal?
- What other abilities does it require?
- Which difficulties are intentional?
- Which difficulties are accidental?
- What plausible wrong route does the task expose?
- Could a shorter task reveal the same mechanism more cleanly?
- When should authentic complexity be restored?
- How will support be faded?
- What will change in the transfer version?
- Will success and failure both tell me something useful?
The best English material is not the one with the most questions. It is the one whose questions make the learner’s next useful decision visible.
That is the Task Laboratory.
Next: The English Tutor Handbook Vol No.0014 | Difficulty Calibration — Making English Hard for the Right Reason.
Explore the English Tutor Handbook. Related teaching guides: Vol. 0014 — Difficulty Calibration; Vol. 0023 — Text Complexity Triage.