Did you know that a careful pupil can lose confidence because of a poorly written multiple-choice question rather than a gap in learning? The child may find two answers that both satisfy the printed instructions and then be told the key recognises only one. That is a question-quality problem worth investigating, not something to dismiss as fussiness.
When a supposedly single-answer multiple-choice item has two defensible answers, check the exact question and every option against the stated conditions. Identify whether the ambiguity comes from missing information, overlapping answers or a vague word such as ‘best’. Then repair the question or answer choices so that the intended learning skill—not the ability to guess the writer’s intention—determines the answer. If the question has already been used, a teacher should review it fairly before drawing conclusions about the pupil’s understanding.
A good question measures an idea, not mind-reading
One-answer questions work when a knowledgeable pupil has a principled reason to select one alternative. If two alternatives satisfy every condition, choosing the official key may become a guessing game about hidden intent. That weakens the value of the item as evidence of learning. Teachers need to know whether students understand the concept, not whether they happen to think exactly like the question writer.
Begin with the intended learning target. Are children identifying multiples, interpreting an experiment, reading a passage or evaluating evidence? Then test the question by temporarily covering the options. Can a pupil understand what kind of answer is requested from the stem alone? If the prompt is unclear before alternatives appear, the choices may be disguising a missing instruction.
Worked example: two correct multiples
Imagine a fictional practice question: ‘Which number is a multiple of 6?’ with options A: 12, B: 18, C: 25 and D: 29. Both 12 and 18 are multiples of 6, because 6 × 2 = 12 and 6 × 3 = 18. If the instructions say ‘choose one answer’, the item has two correct answers as written. Marking A wrong merely because the answer key says B would not measure whether the student understands multiples.
- Learning target: identify multiples of six.
- Check A: 12 ÷ 6 = 2 exactly, so A satisfies the question.
- Check B: 18 ÷ 6 = 3 exactly, so B also satisfies the question.
- Check C and D: neither 25 nor 29 is evenly divisible by six.
- Diagnosis: there are two qualifying choices, not one.
One repair is to change A to 14 while keeping the same stem. Then B = 18 is the only listed multiple of six, and the mathematical concept stays the same. Another is to keep both 12 and 18 but change the instructions to ‘Select all numbers that are multiples of six’, making clear that more than one selection is required. Which repair is appropriate depends on the learning target and assessment format.
A second worked example: dimensions and a missing condition
Suppose a rectangle has area 24 cm² and the task asks, ‘Which dimensions could it have?’ The options include 3 cm × 8 cm and 4 cm × 6 cm. Both are correct because each pair multiplies to 24 cm². The learner is not wrong for noticing that the area alone does not determine one shape.
If the intended skill is using both area and perimeter, revise the stem: ‘A rectangle has area 24 cm² and perimeter 22 cm. Which dimensions could it have?’ The 3 by 8 rectangle has area 24 cm² and perimeter 2(3 + 8) = 22 cm; the 4 by 6 rectangle has the same area but perimeter 2(4 + 6) = 20 cm. With distractors carefully checked, the revised question can genuinely distinguish the intended answer.
Do not silently add a condition to the answer key after pupils have responded. The condition needs to be visible in the question itself. Children should not be expected to infer a measurement, restriction or assumption that the writer omitted.
What about questions asking for the ‘best’ answer?
Some multiple-choice tasks are designed to select a single best response among several partially plausible ones. That can be legitimate when the stem supplies clear criteria and there is a defensible ordering. The word ‘best’ by itself does not solve ambiguity. A question such as ‘Which is the best container?’ is incomplete unless the purpose and relevant constraints are established.
For example, a class needs a container that holds at least 500 mL, is safe to carry for the assigned activity and fits within an explicitly stated size limit. The best option could be determined only after those conditions and the alternatives are given. Without such conditions, two children might reasonably choose differently for different purposes.
The University of Waterloo’s guidance on designing multiple-choice questions encourages a clear single best answer with plausible but distinguishable distractors. This is guidance for question construction, not a claim that any individual school examination has automatically adopted the exact same format.
The difference between two valid answers and two tempting answers
A well-designed distractor can be tempting because it reflects a common error. For example, 18 might be an appropriate correct answer to ‘Which is divisible by six?’ while 19 might be a plausible arithmetic mistake. That does not make 19 a valid answer. The child should explain the rule and test the alternatives, not simply say that two choices ‘look similar’.
Conversely, if an alternative satisfies every mathematical or textual condition, it should not be dismissed merely because it was not intended. This requires careful reasoning. When working with reading passages, the valid answer depends on the wording and support in the passage, not simply on whether the statement could be true in the outside world.
A practical audit in six moves
- Copy the stem exactly and underline words controlling the answer: one, all, except, most likely, according to the passage, or under the stated conditions.
- State the learning target in plain language. What understanding should the item reveal?
- Cover the answer choices and decide what a suitable answer would need to satisfy.
- Test each option against the exact stem, writing a short justification for accepting or rejecting it.
- If two options satisfy the criteria, locate the missing distinction or overlapping options rather than guessing which was intended.
- Revise the stem or distractors, ask another reader to solve the revised version independently and record the agreed answer.
Teachers can run this audit before a quiz is distributed. When time is limited, prioritise questions with vague qualifiers, multiple numerical solutions or large differences between the question’s stated goal and the answer key’s explanation. The purpose is not to make an easy item harder; it is to remove avoidable ambiguity while retaining a meaningful learning check.
A student’s response when two answers seem possible
First, reread the full question. Is it asking for the only correct answer, all correct answers, the most supported answer, or an exception? Check for a condition the child may have missed. If two choices still seem valid, write a brief explanation for each. In ordinary homework or practice, show the reasoning politely to the teacher.
During a formal examination, pupils must follow the examination instructions and applicable rules; this article does not recommend disputing an item while the test is underway or assuming every apparent ambiguity will earn credit. A later teacher discussion can separate a genuine item problem from a missed instruction or conceptual misunderstanding.
How teachers can respond fairly after a flawed item was used
A teacher may have several reasonable options for classroom formative practice: clarify why one response is intended, accept multiple defensible answers if appropriate, revise the item and reteach the target concept, or exclude the flawed item from a simple classroom interpretation of progress. The right procedure depends on the assessment’s status, school rules and fairness to all students. Do not promise that a particular mark adjustment is automatic.
The University of Michigan’s guidance on designing effective examinations stresses matching items to learning outcomes and using clear stems with a defensible correct or best response. The broader lesson is that assessment design and the interpretation of scores are connected.
A parent can help without turning a mistake into a conflict
Ask the child to show the exact printed question and explain why both alternatives qualify. Then test the reasoning together. If the second option is valid only after adding an imaginary condition, explain that distinction kindly. If both really qualify as written, encourage a concise, respectful question to the teacher: ‘Could you explain which condition makes B the only answer? I found that A also satisfies the stated rule.’
Keep the focus on learning. A disagreement over a single practice question need not become a dispute about a teacher’s competence or a child’s intelligence. The best outcome is a clearer question, a more accurate understanding of the topic, and a fairer account of what the pupil can do.
The progressive learning route for different ages
Younger learners can begin with simple number and picture choices that clearly have one answer. Next introduce a deliberately flawed example with two qualifying options and ask them to spot the problem. Older learners can design a question for a partner, check its alternatives, revise it and test whether the partner can explain the correct choice without guessing.
For advanced students, discuss that ‘one best answer’ requires a defined decision rule. A scientific question may depend on what evidence is provided and what uncertainty is acceptable; a mathematical item may depend on domain restrictions; a reading question may depend on what the passage actually supports. The important transfer skill is identifying the condition that makes the decision defensible.
An independent exit task with answer
Write this item on a card: ‘Which fraction equals one half?’ Options are A: 2/4, B: 3/6, C: 2/5 and D: 3/5. For a single-answer format, the item is flawed: both 2/4 and 3/6 simplify to 1/2. A useful repair is to replace B with 3/8 or change the instruction to ‘Choose all equivalent fractions’. Have the pupil explain why that repair changes the item’s validity without changing the concept of equivalent fractions.
A changed-context transfer prompt: ‘Which day had the highest attendance?’ with two days tied at 28 pupils. If the question asks for one day, a pupil who notices the tie has found a genuine issue. Fix the choices or ask which days shared the highest count. This checks whether the learner can detect the same structural problem outside arithmetic multiples.
Common mistakes when fixing test items
- Changing the answer key without checking whether another option remains defensible.
- Making a distractor absurdly implausible so pupils can answer without the target knowledge.
- Adding a hidden condition to an oral explanation but not to the printed question.
- Treating a partially true alternative as the best answer without identifying the stated criteria.
- Using ‘all of the above’ to hide an unclear or poorly differentiated question.
- Assuming a question was flawed before checking whether the student missed a negative, time period or other qualifier.
Each error has a different repair. A validity problem in the stem requires a rewritten stem. Overlapping alternatives require revised choices. A pupil who ignored a stated qualifier may need comprehension practice rather than a new answer key. Good diagnosis prevents a one-size-fits-all correction.
Limitations of this guide
This article concerns classroom learning and general item-writing principles. Real examinations can use specialised formats, complex scoring schemes and review procedures that differ from everyday homework. Follow the official instructions for the specific assessment. Do not infer from one disputed practice item that an entire examination is invalid or that a student’s grade must change.
Likewise, an independent student explanation is evidence worth reviewing, not proof that an item is flawed. Teachers should consider accurate subject knowledge, the intended level, the full question context and the applicable marking scheme before deciding.
Frequently asked questions
Are all multiple-choice questions supposed to have exactly one correct answer?
No. Some explicitly allow multiple selections, and some ask for one best response. The instructions and scoring format must make the intended task clear.
What if two answers are true but one does not answer the question?
A true statement is not necessarily a correct response to that stem. Relevance to the requested property or evidence is essential.
Should a parent tell the child always to challenge the key?
No. Encourage checking the instructions and explaining the reasoning respectfully. The goal is understanding, not routine confrontation.
Can a flawed question still teach something useful?
Yes, when it is recognised and discussed. Analysing why a question fails can clarify a mathematical condition or the limits of a reading claim.
Can an item be fixed without becoming easier?
Often. Replace overlapping alternatives or state a missing condition while preserving the intended intellectual challenge.
Continue through eduKate
The Core Aim guide to fair marking of differently worded short answers examines a related but distinct issue: fair evaluation of an answer after a clear question has been posed. This article focuses on the quality of the question itself. See also the eduKate Education learning library and the Clementi Mathematics teaching reference for explanation-first learning and independent checking.
Good assessment should invite a learner to show what they know. A carefully checked question gives the child a fair chance to do that, and gives the teacher evidence worth trusting.