Multiple Choice Distractor Design: How to Write Wrong Answers That Work

TL;DR. A multiple-choice question is only as good as its wrong answers. Plausible distractors test knowledge; obvious distractors test elimination. Three patterns produce strong distractors: common student misconceptions, partially correct statements, and confusable adjacent concepts. Random wrong answers are the most common — and most damaging — mistake.

Why distractors matter more than the right answer

In a four-choice question, the test-taker has a 25% chance of guessing correctly. That floor moves up dramatically based on distractor quality:

A question with three obvious distractors tests *recognition of obviously wrong answers*, not knowledge of the right one. It's a question that grades nothing.

The rule of thumb in assessment design: spend more time writing your distractors than your correct answer.

Three distractor patterns that work

Pattern 1 — Common student misconceptions

The strongest distractor is the answer a student *would* give if they confused two related concepts.

Topic: photosynthesis.

Question: "What gas do plants release during photosynthesis?"

Weak distractors: nitrogen, hydrogen, methane (random gases).

Strong distractors: carbon dioxide (the *input*, commonly confused with output), water vapor (a related but distinct exchange), nitrogen (a real plant intake confusing with photosynthesis).

The student who picks "carbon dioxide" reveals a specific misconception (confusing input with output). The teacher learns something useful from the wrong answer. This is what good assessment does.

To find common misconceptions: ask the topic teacher "what do students get wrong here, and why?". The answers become your distractors.

Pattern 2 — Partially correct statements

A distractor that's *true* but doesn't fully answer the question.

Topic: causes of World War I.

Question: "Which factor was the most immediate cause of World War I?"

Weak distractor: "the invention of the printing press" (obviously wrong, off-topic).

Strong distractor: "the rise of nationalism in Europe" (true and contributing, but not the *most immediate* cause; the assassination of Archduke Franz Ferdinand was).

The student must distinguish between contributing factors and the precipitating event — Bloom's Analyze level. The question now tests reasoning, not just memory.

Pattern 3 — Confusable adjacent concepts

In any field, certain concept pairs get confused. Use both as choices.

If your topic has a known confusion pair, the wrong half is your strongest distractor.

Distractor antipatterns

The mistakes that destroy multiple choice questions:

"All of the above" / "None of the above"

These are not distractors; they're escape hatches. Students who don't know learn to pick "all of the above" for safety, and the question loses its discrimination. Skip them entirely.

Length tells

Students learn that the longest answer is usually the correct one (because writers add qualifiers to avoid technical errors). If your correct answer is twice as long as your distractors, you're giving away the answer.

Fix: make all four choices roughly the same length, or vary length across questions so length doesn't correlate with correctness.

Grammatical tells

The stem and the correct answer must agree grammatically. If three distractors are noun phrases and one (the correct one) is a clause, students notice.

Repetition of stem keywords

If the stem says "the muscle responsible for…" and the correct answer is "the cardiac muscle", that's a tell. The word "muscle" appearing in both flags the answer.

Fix: rewrite the correct answer so it doesn't echo stem keywords. Echo them in distractors instead.

Random wrong answers

The single biggest sin. "What's the capital of France? A) Paris B) Wednesday C) Mitochondrion D) 47" tests nothing. Distractors must come from the same conceptual neighborhood as the correct answer.

A workflow for writing distractors

Start by writing the correct answer. Then ask three questions:

If you can answer all three, you have a strong question. If you can't, the question may not be testing what you think it's testing.

For more on the broader question-writing process, see How to Write Hard Quiz Questions.

How AI handles distractors

Out of the box, AI quiz generators default to weak distractors — random wrong answers that share keywords with the topic but no conceptual relationship.

The fix is the prompt:

"For each multiple-choice question, the three distractors must each represent one of: (a) a common student misconception about this topic, (b) a partially-correct statement that doesn't fully answer the question, or (c) a confusable adjacent concept. Do not include random wrong answers."

With this instruction, modern LLMs produce distractor quality close to a human expert. Without it, they don't. See AI Quiz Generator Explained for more on prompt-driven quality.

How to evaluate your existing distractors

Take any multiple-choice quiz you've made. For each question:

Run this check on 10 questions and you'll spot the patterns in your own writing.

Frequently Asked Questions

How many distractors should a multiple-choice question have?

Three (four total choices) is standard. Two distractors (three choices) is acceptable for younger students or when you can't write three plausible options. More than four creates fatigue without adding meaningful discrimination.

Should distractors be the same length as the correct answer?

Yes. Length is the most common giveaway. Aim for choices that are within ±20% of each other in word count.

What if I can only think of two good distractors?

Make it a 3-option question. Better to have three strong choices than four with one obvious filler.

Can distractors include partial truths?

Yes — see Pattern 2 above. The key is they must be *insufficient* answers, not *correct* ones. An answer that's defensible by some standard isn't a distractor; it's a flaw.

How do I tell if my distractors are working?

Item analysis. After 30+ students take the quiz, look at which wrong answer was picked most often. If 60% of wrong answers cluster on one distractor, it's a strong distractor. If wrong answers spread evenly across all three, two of them are too obvious.

---

Want quizzes with AI-generated, misconception-based distractors out of the box? Try SimpleQuizMaker free. Back to the How to Make a Quiz pillar guide.

A quick decision framework for writing distractors under time pressure

Most teachers don't have 20 minutes per question to craft the perfect wrong answer. Here's a faster path that still avoids the "obviously wrong" trap.

Common mistakes that undermine otherwise good distractors

If you're building quizzes by hand, running these checks question-by-question adds up fast. Tools like SimpleQuizMaker's AI quiz generator are trained to lean on the misconception and partial-truth patterns above by default, which is a useful starting draft even if you plan to hand-edit before publishing — free accounts get 5 AI generations a month to test the workflow, with paid plans (see pricing) raising that ceiling for classroom-scale use.

Worked example: turning a weak item into a strong one

Weak version — "What is the capital of France?" A) Paris B) A car C) A sandwich D) Blue. Every wrong answer is a different category entirely; a student who has never heard of France still scores 100% by elimination.

Strengthened version — "What is the capital of France?" A) Paris B) Lyon C) Marseille D) Nice. Now every distractor is a real French city, so guessing correctly requires actually knowing which one is the capital, not just knowing the topic is geography. This is the partially-correct-category pattern applied to a simple factual question, and it works the same way whether you're writing for test-maker style summative exams or a quick formative check.

For teachers building this into a full workflow — importing existing material as source content rather than writing from scratch — creating a quiz from a PDF or slide deck and then hand-tuning the generated distractors against the patterns above is usually faster than writing every item cold.