How to design eLearning quizzes that measure learning

A quiz is useful only when its questions provide evidence for a learning outcome. Before writing items, identify the knowledge or decision the learner must demonstrate and the level of performance required.

If an outcome says “choose the correct escalation route,” a question about the definition of escalation is weak evidence. Ask the learner to choose a route in a realistic case.

List each outcome, its importance, the type of evidence required and the number of questions or tasks that will sample it. This prevents a ten-question quiz from accidentally testing one easy topic repeatedly while ignoring a critical outcome.

Weight questions according to risk and importance rather than how many slides a topic occupied.

Put the problem in the question stem so options can stay concise. Give all information needed to answer, but remove irrelevant detail unless interpreting that detail is itself part of the skill.

Avoid trick wording, unnecessary negatives and clues created by grammar or option length. The challenge should come from the subject matter.

Wrong answers should represent believable misconceptions, shortcuts or decisions. Ask SMEs what people commonly get wrong and why. Those errors are far more useful than obviously silly distractors.

Keep options parallel in style and specificity. If one option is much longer and more qualified than the others, it can reveal the answer without testing knowledge.

For judgement, customer service, safety, leadership or compliance decisions, give enough context to make the choice meaningful. Ask what the learner should do next, what risk matters most or which response best follows the policy.

Do not bury a simple recall question inside a paragraph of fictional context. A scenario earns its place when the context changes the reasoning.

“Correct” and “Incorrect” confirm a result but teach very little. Explain the principle behind the correct response and, where useful, why a tempting alternative fails.

For practice questions, immediate explanatory feedback is valuable. For high-stakes final assessments, consider whether revealing answers after each attempt undermines the integrity of retakes.

Decide the pass mark based on the level of evidence required, not because 80% happens to be familiar. Critical safety or regulatory decisions may need stronger rules than low-risk awareness content.

Make sure unanswered graded questions remain part of the possible score. Define retakes, randomisation and whether learners can review previous responses.

• Can the question be answered without reading the course?
• Does it map to a stated outcome?
• Is there one defensible best answer?
• Are distractors plausible but clearly wrong under the stated conditions?
• Can keyboard and assistive-technology users operate it?
• Does feedback explain the reasoning?
• Does scoring behave correctly when a question is skipped?

After launch, review which questions everyone misses or everyone passes. That pattern may reveal a bad item rather than a learner problem.

For alignment from the start, read writing outcomes and assessments.