Writing Quiz Questions That Teach Instead of Just Testing

Most quizzes in online courses test whether someone was awake. They are easy to write, easy to pass, and teach nothing.

A well-written question does something different: it makes the learner apply what they just learned, and shows them precisely what they missed.

Start from what they should be able to do

• Write the capability first, then the question that would prove it.
• Ask about applying, not recalling — recall questions mostly test recent memory, which fades.
• Use a realistic situation. Give a scenario and ask what the learner would do.
• If a question can be answered without taking the lesson, it is not testing your lesson.

Question types and what they are good for

Multiple choice works well for judgement between plausible options. It works badly when the wrong answers are obviously wrong.

Multiple response — select all that apply — is harder and better at catching partial understanding, but tell learners how it is scored.

Short written answer reveals real understanding but needs marking, so use it sparingly and at key points.

Ordering or matching works for processes and relationships.

True or false is the weakest type — a coin flip gives fifty percent. Use it rarely, and only for genuinely common misconceptions.

Writing the options

This is where most quizzes fall apart. The wrong answers matter more than the right one:

• Make every wrong option plausible. Each one should represent a real mistake someone might make.
• Base distractors on actual misconceptions you have seen learners have — these are the most valuable questions you can write.
• Keep options similar in length. The longest option being correct is a pattern learners exploit without even noticing.
• Avoid all of the above and none of the above — they usually test logic rather than knowledge.
• Avoid absolutes like always and never in wrong options; experienced test-takers eliminate them automatically.
• Keep the grammar consistent between the question and every option, or the grammar gives the answer away.

Feedback is where the learning happens

The single highest-value change you can make to a quiz:

• Explain why the correct answer is correct — not just that it is.
• Explain why each wrong answer is wrong. This is where a distractor based on a real misconception pays off.
• Point back to the relevant lesson so they can review the specific part.
• Show feedback immediately for practice quizzes — delayed feedback is much weaker.

A quiz with good feedback is a teaching tool. A quiz with only a score is a filter.

Practice quizzes versus final assessment

• Practice quizzes: unlimited attempts, immediate detailed feedback, no stakes. Their purpose is learning, and they work best placed right after the relevant lesson.
• Final assessment: limited attempts, covers the whole course, feedback after submission. Its purpose is confirming capability.
• Do not mix the two. A high-stakes quiz early makes people cautious rather than exploratory.

Common mistakes

• Questions that only test recall of the last five minutes.
• Trick questions — they test reading care, not understanding, and they irritate people.
• Ambiguous wording where two answers are defensible.
• Testing content you never taught.
• Too many questions at once — short and frequent beats one long quiz.
• No feedback at all — the most common and most wasteful mistake.

Improving quizzes with data

If your platform shows question-level results:

• A question almost everyone gets right is probably too easy to be worth asking.
• A question almost everyone gets wrong usually means the lesson was unclear or the question is ambiguous — check the lesson first.
• Look at which wrong option people choose — a popular distractor points to a specific misconception you can address directly.
• Revise a few questions each time you update rather than rewriting everything.

Accessibility and fairness

• Keep language plain — you are testing subject knowledge, not reading speed.
• Be careful with time limits; they disadvantage non-native speakers and people using assistive technology without testing anything you meant to test.
• Avoid culture-specific examples that some learners will not recognise.
• Do not rely on colour or images alone to convey a question.

One thing worth remembering

Ask one question about every quiz item you write: if a learner gets this wrong, will they know what to do about it?

If the answer is no, the question is measuring something rather than teaching it — and in a course, teaching is what you were hired for.

FAQ