A Secondary Mathematics question contains a paragraph, two prices, a percentage change and a final instruction to justify whether a recommendation is reasonable. The student notices the percentage sign, immediately calculates something and discovers that the number obtained does not answer the question. In English, another learner writes an excellent description but the examination question asked for both an explanation and an evaluation.
These errors can occur even when the underlying subject knowledge is available. The first failure happens before calculation or writing: the student has not unpacked what the question actually requires. Question decomposition is the skill of separating a complex task into its deliverables, givens, unknowns, constraints, relationships and checks so that the learner can decide what to do without losing the overall purpose.
Decomposition does not mean mechanically splitting every sentence into a checklist. It is a method for making a demanding question manageable without breaking its meaning. Some tasks have parts that can be solved independently; others depend on a condition established earlier. A sound decomposition shows which pieces connect and which actions must be completed before a defensible final answer is possible.
Correct decomposition can help students understand word problems, scientific investigations, source-based questions and essay prompts. But it cannot replace missing subject knowledge. If the learner does not understand what a gradient represents or how to justify an inference, a beautifully annotated instruction sheet may still lead to an incorrect answer.
This article connects with Problem-Framing Skills (choosing the right problem), Sequencing Skills (ordering actions), Annotation Skills (marking the evidence in source material), Practice Design Skills (constructing learning tasks) and Exemplar Analysis Skills (learning from a model). Here the owned job is the transition from a *given question* to a structured response plan.
Before the Top 10: Read for the Finish Line
The first question to answer is surprisingly simple: What does the examiner, teacher or real-world task actually want at the end? A numerical value, a comparison, a labelled diagram, an explanation, an evaluation or a reasoned recommendation are different deliverables. A task can legitimately ask for several; the student must not supply one while overlooking another.
Then distinguish information from instructions. A diagram or passage may contain many details, but only some are relevant to the current answer. The question may also specify conditions that cannot be ignored, such as an allowed method, a time interval, a comparison basis, a prescribed audience or a request to justify a conclusion using the data.
Decomposition is most useful when it preserves dependencies. A student asked to compare two rates must first identify the rates and their units. A writer asked to evaluate an argument must first understand what the argument claims and how its evidence is meant to support it. A learner who jumps directly to evaluation can end up criticising a position the source never asserted.
The ten headings are an eduKate / Wintour House teaching synthesis, not a universal official MOE/SEAB ten-step examination routine. They draw on problem-solving and metacognitive teaching principles, but should be adapted to task complexity and individual learner needs.
FINAL DELIVERABLE → GIVENS → CONDITIONS → MODEL AND STEPS → CHECK AGAINST THE ORIGINAL QUESTION
1. Identify the Required Final Product Before Doing Any Working
A question can contain an attractive number, interesting story or familiar method while asking for a different final output. Begin by naming the deliverable in plain language: find a quantity, make a decision, explain a pattern, compare two cases, justify a method or evaluate a source.
A Mathematics problem may contain discount percentages but conclude, “Explain whether the cheaper advertised option remains cheaper after delivery costs.” The task is a comparison and explanation, not merely a discount calculation. An English question may say “Describe how the writer develops a sense of tension,” which requires attention to development across the passage rather than listing one language device.
Write a short finish-line note: “I must return a comparison with a justified conclusion,” or “I must explain the relationship shown in the graph.” After completing working, compare the result against that note rather than assuming that any accurate calculation is enough.
Worth learning because naming the required product keeps the learner’s effort aligned with the actual assessment demand.
2. Separate Given Information, Unknowns and Missing Data
The distinction between what is supplied, what must be found and what remains genuinely unavailable is foundational. Students often fill an information gap with an unspoken assumption because the task resembles a familiar example. Good decomposition makes those categories visible.
A word problem gives two totals for different combinations of items and asks for individual prices. Mark the totals as given and the individual prices as unknowns. If only one independent total is available for two unknown prices, there may not be enough information for a unique solution. A Science report that lists outcomes but no control conditions may allow a description without supporting a strong causal explanation.
Use a small three-column note: GIVEN, REQUIRED, NOT YET KNOWN. Do not promote an inferred value into the first column without checking. If a missing datum is truly necessary, state the limitation or identify the additional information needed.
Worth learning because a transparent information boundary prevents students from inventing facts to make a problem look solvable.
3. Mark Conditions and Restrictions Before Applying a Familiar Rule
A problem may contain details that limit which formula, interpretation or writing structure is appropriate. Domain restrictions, units, variable conditions, prescribed audience and evidence limits can change the answer. Students who treat conditions as decorative prose often choose a familiar method that is invalid here.
In Mathematics, dividing by an expression involving x may require checking whether that expression could equal zero. In Science, a conclusion from a steady simple series circuit cannot automatically be transferred to a parallel network. In English situational writing, the intended recipient changes tone, level of detail and appropriate call to action.
Circle every statement that restricts the response: “assuming,” “for this period,” “using information from the text,” “justify,” “at most,” or “to the nearest.” Explain why each condition matters. A restriction that changes no possible part of the solution may be background; one that changes method validity must remain visible.
Worth learning because conditions determine where an otherwise correct method can and cannot be used.
4. Translate Dense Words Into a Clear Task Model
Some word problems overload readers with context before revealing the underlying structure. Translation may use a bar model, table, labelled sketch, equation, timeline or short prose restatement. The new representation must preserve quantities and relationships; it is not an excuse to replace the actual question with an easier one.
A Primary problem states: “A container holds 18 blue and red counters. There are six more blue counters than red counters.” Let red counters be r, then blue counters are r + 6 and 2r + 6 = 18. Solving gives r = 6 and blue = 12. A bar model with equal base units and an extra six shows the same relationship without changing it.
After drawing or writing the model, point from each important element back to the original sentence. If a bar segment or equation term has no support in the question, remove it. If a condition from the question is missing, repair the model before calculating.
Worth learning because an accurate intermediate representation frees attention for reasoning while preserving the problem’s meaning.
5. Distinguish Subtasks That Can Be Solved Independently From Those That Depend on Others
Some long questions can be divided into independent pieces; others form a chain. For example, finding the perimeter and finding the area of a rectangle are different calculations from the same dimensions. But computing the cost of flooring may depend on the area already being established.
In a Science investigation, identifying the independent variable may not depend on computing a mean, while explaining the reliability of the conclusion may depend on sample design and the observed variation. In English, evaluating an argument’s strength depends on first identifying the author’s actual claim and the supporting evidence.
Draw a simple dependency chain when needed. Mark which step supplies a value or premise used later. Do not reverse the sequence by using a later conclusion as though it were an earlier fact. Avoid making the chain longer than the task requires.
Worth learning because correct dependencies reduce repeated work and prevent circular explanations.
6. Recognise When a Task Contains More Than One Command Word
A question may ask students to identify and explain, compare and evaluate, or calculate and justify. Supplying only the first command can produce an incomplete response even when that portion is accurate. The commands are not synonyms; each imposes a different evidence or reasoning requirement.
In a Science question, “State one observation and explain its significance” asks for a directly reported outcome and then a reasoned interpretation. In English, “Compare how two writers present a place and evaluate which description is more effective for its purpose” requires a same-basis comparison followed by criteria-based judgment. Listing techniques is not enough.
Underline command verbs and write a mini-deliverable for each. When the task says “using evidence,” include a place for that support. When it says “evaluate,” identify what standard of effectiveness will be used rather than substituting personal preference for analysis.
Worth learning because command-word decomposition protects against answers that are polished but incomplete.
7. Choose Methods Only After the Structure Is Visible
Students sometimes select a formula because a familiar number appears. Decomposition should delay method choice until the variables, relationships and required result are clear. The same quantities can occur in percentage change, direct proportion and fixed-fee models, which require different treatments.
Compare two pricing plans. Plan A charges $8 plus $2 per session; Plan B charges $3 per session, with no fixed fee. At ten sessions, A costs 8 + 2(10) = $28 while B costs 3(10) = $30, so A is $2 cheaper. At five sessions, A costs $18 and B $15, so B is $3 cheaper. One blanket statement about the cheaper plan would be wrong without specifying usage.
The correct decomposition identifies a fixed component, a variable rate, the number of sessions and the comparison target. A graph or equations may help. The check is whether the method represents the actual costs, not whether the student used a formula that looked sophisticated.
Worth learning because method choice should follow problem structure rather than superficial resemblance to a memorised worksheet.
8. Allocate Answer Space and Attention According to the Real Demands
Decomposition should improve efficiency, not consume the entire examination period. A simple factual item may need a direct response. A multi-step argument or modelling task deserves more planning. Students should judge which parts are essential and where their time will produce the most informative work.
For an extended English response, the student can note claim, evidence, reasoning and qualification before composing paragraphs. For a Mathematics problem, they may first mark the givens, show the model and reserve a final line for the requested decision. The plan need not become a separate essay inside the answer booklet.
Use a proportional approach: brief check for routine items, structured outline for complex tasks, and a visible final-answer check where omission is costly. Avoid interpreting all examination difficulty as a problem of speed; some pupils need conceptual instruction rather than faster annotation.
Worth learning because task structure helps learners spend effort on the work that the question actually assesses.
9. Recombine the Pieces Into One Coherent Answer
A question can be correctly decomposed yet answered as a disconnected list. The final response must show how the subtasks relate to the main demand. A set of calculations is incomplete if the problem asked whether an option is reasonable; a list of quotations is incomplete if the question required an explanation of effect.
In the pricing example, the final conclusion should state which plan is cheaper at the specified number of sessions and support it with the computed totals. In Science, the conclusion should connect observations to the mechanism only as far as the evidence warrants. In English, the response should use comparisons to answer the overall question rather than leaving isolated feature lists.
Read the original question once more after drafting. For each command and condition, locate the part of the response that satisfies it. If a sub-answer is correct but irrelevant, remove or reposition it. If a required justification is missing, add the reasoning rather than decorating the existing sentences.
Worth learning because decomposition is only successful when the pieces return to the original problem as a justified whole.
10. Check the Answer Against the Original Task and a Boundary Case
A calculated result can be arithmetically correct yet wrong for the question. A written argument can be fluent yet answer a different prompt. The final check returns to the source task, units, constraints and expected direction. Where useful, a boundary or simple test case can expose an invalid interpretation.
For a counter problem, verify that the red and blue counts total 18 and differ by six: 6 + 12 = 18 and 12 – 6 = 6. For the pricing problem, substitute the given number of sessions into both cost expressions and check which total is smaller. In a source-based response, verify that the quotation actually supports the stated inference and is not taken out of context.
Record recurring failures by stage: missed command, wrong model, incorrect dependency, calculation slip or conclusion beyond evidence. Practise the first consistently weak stage with a fresh task. Do not automatically repeat the entire topic when the issue is an omitted interpretation step.
Worth learning because checking against the original task distinguishes genuine problem-solving from reaching a plausible but irrelevant result.
The Top 10 Question Decomposition Skills as One System
The sequence is DELIVERABLE → GIVEN/UNKNOWN → CONDITIONS → MODEL → DEPENDENCIES → COMMAND WORDS → METHOD → EFFORT → RECOMBINE → CHECK. A pupil may run through it mentally in seconds for a familiar problem. A complex unseen task may require an explicit sketch or short written outline.
The practical question is: “What must my final answer contain, and what information and reasoning must I establish to reach it?” Decomposition should reduce cognitive load, not create another assignment on top of the original one.
Question Decomposition Is Not Problem Framing or Sequencing Alone
Problem-Framing Skills owns deciding what problem to solve in an open situation. Question decomposition starts once a specific question has been given and the learner must interpret its demands accurately.
Sequencing Skills owns arranging steps or events in a valid order. Decomposition includes sequencing but also identifies givens, constraints, command words and the final required product.
A Worked Primary Mathematics Decomposition
Question: “There are 18 red and blue counters altogether. There are six more blue counters than red counters. How many of each colour are there?” First, final products: red count and blue count. Givens: total 18; blue exceeds red by six. Unknowns: two counts. Relationship: if red = r, blue = r + 6.
Model: r + r + 6 = 18. Hence 2r = 12 and r = 6; blue = 12. Verification: 6 + 12 = 18 and 12 − 6 = 6. The answer is six red counters and twelve blue counters.
A bar model could show two equal groups plus six extra, but its labels must preserve the same relationship. The pupil should then solve a similar changed question without copying the worked representation.
A Worked Secondary Mathematics Decomposition: Two Pricing Plans
Plan A charges a fixed $8 plus $2 per session. Plan B charges $3 per session with no fixed fee. The question asks which is cheaper at ten sessions and why. First identify the required output: a comparison and explanation at ten sessions. Then distinguish fixed fee, variable rates and usage count.
Form A(n) = 8 + 2n and B(n) = 3n. At n = 10, A(10) = $28 and B(10) = $30. Therefore Plan A is $2 cheaper at ten sessions. A(n) = B(n) when 8 + 2n = 3n, so they are equal at n = 8. For fewer than eight sessions, B is cheaper; for more than eight, A is cheaper, assuming nonnegative session counts and unchanged terms.
The boundary case n = 8 shows why “Plan A is always cheaper” would be an unjustified generalisation. The question must be decomposed by use case before drawing a conclusion.
A Worked English Comprehension Decomposition
Prompt: “Compare how the two passages create tension and explain which method is more effective for the intended reader.” Deliverables: identify a technique and evidence from each passage, compare on an appropriate basis, and make a criteria-based judgment suited to the audience.
A weak response lists imagery in Passage A and punctuation in Passage B with no explanation. A stronger plan asks how each feature shapes anticipation, pace or uncertainty in context. The final judgment should explain its criterion rather than announce “I prefer Passage A.”
The purpose is not to impose one fixed literary verdict. The student must justify the comparison using relevant words and the task’s actual demand.
A Worked Science Decomposition: Investigating Dissolving
Prompt: design a fair comparison of how water temperature affects the time a fixed amount of a particular solute takes to dissolve, under a specified stirring arrangement. Required output: identify the independent variable, dependent measure and relevant controls, then explain why those controls matter.
Independent variable: water temperature. Dependent measure: measured time taken to dissolve under clearly defined conditions. Controls might include amount and form of solute, volume of water, container and stirring method. The selected values and safe laboratory procedures should be teacher-approved.
The explanation should not assert a universal causal law from one trial. Appropriate repeat measurements and fair comparisons improve what can be learned, while the experiment’s design and actual results determine the conclusion.
For Primary Students: Name the Question’s Finish Line
Primary learners can begin by underlining what is being asked and circling relevant quantities or evidence. Teachers can model restating a word problem and building a simple bar model. If a child cannot interpret the language, help with vocabulary before judging mathematical understanding.
A reliable short routine is: what do I know; what must I find; what relationship connects them; does my answer fit? The words should remain simple and adaptable.
For Secondary Students: Catch Compound Demands
Secondary learners meet longer questions with mixed concepts. Practise interpreting conditions and multiple command words before computing or drafting. In Mathematics, identify domain and units; in Science, separate observation and explanation; in English and Humanities, distinguish identify, compare, infer and evaluate.
Mixed practice should eventually remove the chapter heading so learners must select the method. If task decomposition is accurate but performance still fails, investigate subject knowledge rather than adding ever longer checklists.
For JC Students: Preserve the Assumptions
In JC Mathematics and Science, a task’s hidden complexity may involve assumptions, model validity or several dependent calculations. Students should state key conditions before rearranging equations and distinguish empirical evidence from theoretical inference. In General Paper, they should identify the precise contention and which evidence standard the evaluation requires.
A useful outline separates claims, premises and limitations without writing an entire essay before the answer. The aim is reasoning economy rather than procedural bureaucracy.
A Fifteen-Minute Question Decomposition Clinic
Minutes 1–3: read the task and state the final product. Minutes 4–6: mark givens, unknowns, conditions and command words. Minutes 7–9: choose a model or method with a reason. Minutes 10–12: work the problem or outline the answer. Minutes 13–15: recombine, check units and evidence, and identify one boundary case.
This is an illustrative classroom sequence, not a researched optimum. Straightforward tasks may take seconds; complex investigations require longer planning and subject-specific teaching.
Question Decomposition With AI Assistance
AI can turn a long prompt into subtasks, but it can also alter the question, introduce assumptions or reveal the solution during supposed planning. If a student uses an AI tutor, it can be instructed to identify the deliverables and givens without supplying the calculation or complete answer.
Compare the AI outline against the original prompt word by word. Check whether any requirement disappeared or any invented constraint appeared. Then close the scaffold and solve a fresh suitable task independently. Avoid uploading identifying pupil records or private school materials without appropriate rights and permission.
The learner should understand the breakdown, not merely possess a convenient decomposition prepared by another system.
The Wintour House Test: What If AI Can Solve Every Question?
If a model returns polished solutions immediately, an independent learner still needs to know whether the solution answers the actual question. A fluent answer can miss a command word, violate a condition or compare the wrong quantities. Someone must be able to audit that alignment.
The Gold Standard of decomposition is a controlled chain from prompt to result: every important demand preserved, every method justified and the final answer checked against the original task.
Research Anchors and Limits
The Education Endowment Foundation’s Improving Mathematics in Key Stages 2 and 3 (2017) recommends teaching pupils to use and compare problem-solving approaches and to reflect on methods. It is guidance for a UK educational context, not a list of compulsory Singapore examination annotation steps.
EEF’s discussion of children’s mathematical problem-solving models the REEL prompts—Read, Explore, Exemplify and Link—as one classroom aid. The article also emphasises teacher think-alouds and the plan–monitor–evaluate cycle.
EEF’s problem-solving reflection cautions against treating challenging mathematical problems as tasks with one universally rehearsed procedure. This supports teaching method selection and reasoning, not a promise that every problem can be solved with a single checklist.
The EEF’s worked-example guidance explains how comparing strategies and examining reasons can support problem-solving instruction. Worked examples are a support stage; learners still require opportunities to make choices unaided.
The ten headings, pricing problems, argument prompts and classroom routines are Wintour House editorial educational synthesis. They are not official MOE/SEAB mark schemes or guarantees of examination gains. Their purpose is to help learners preserve the demands of a question and show their reasoning in an independent response.
Final question-decomposition test: Can the learner identify all deliverables, preserve given conditions, choose a justified sequence, recombine the pieces and show how the response actually answers the original question?
