Quizzes.
A quiz is the assembled experience of multiple questions — the wrapping that holds your assessment together. A quiz can be great or it can be performative. Most aren't designed; they're assembled. The difference is in the structure, not the questions.
Measuring or reinforcing learning.
Quizzes serve one of two purposes. Either they reinforce learning through retrieval practice (formative), or they measure whether the learner has achieved the objectives (summative). Both are valuable. The trouble starts when courses use one when they meant the other.
- Formative — knowledge checks inside the course, low stakes, high feedback.
- Summative — end-of-module assessment, defined pass mark, less feedback during.
- Pre-course — gauging existing knowledge to adapt the path forward.
- Compliance certification — proof that the learner met the standard.
Filling time or proving you covered content.
A quiz that exists just because "every module needs a quiz" usually tests trivia, not understanding. If the questions don't connect to the objectives, the quiz fails the learner and reveals that the module is unclear about what it actually teaches.
- Trivia about content you didn't intend to assess.
- Self-grading reflection — that's not a quiz, that's a journal.
- Quiz-as-punishment for not paying attention.
- Box-ticking exercises with no consequence.
Take the quiz.
A short formative quiz on quiz design. Three questions, mixed formats (one single-answer, one multi-select, one true/false). Feedback after each. Score at the end. Try it — and notice the design choices as you go.
Loading question…
Quiz complete
Submit is disabled until the learner has selected something. After submitting, the answer locks and the feedback appears — they can't change their mind, but they get to see why. The Next button only appears after feedback.
Get the structure right.
One question per screen, by default
Scrolling quizzes invite skim-and-skip behaviour. One question per screen forces the learner to engage with each question on its own terms. The exception is short quizzes (3-5 questions) where the entire quiz fits on a single screen without scrolling.
Progress indicator
"Question 3 of 8" tells the learner where they are. Without it, every question feels like it could be the last — and the uncertainty is unkind. A subtle counter or progress bar near the top of the question is enough.
Lock answers after submission
Once an answer is submitted and feedback shown, the answer should be uneditable. Otherwise learners change their answer after seeing the feedback — and the quiz becomes a guess-and-correct exercise rather than an assessment.
Accessibility — use native form controls
Build each question on real <input type="radio"> or <input type="checkbox"> inputs. See the radio buttons and checkboxes guides for why. Custom-built quiz UI almost always breaks keyboard and screen-reader behaviour.
Accessibility — announce question changes
When the next question loads, screen readers need to know. Use an aria-live region for the question text, or move keyboard focus to the new question's heading. Without this signal, the page silently changes and screen reader users miss the cue.
Show feedback in context
Per-question feedback should appear directly below the answered question, not in a modal or a separate screen. The learner needs to see their answer, the result, and the explanation together. Modal feedback breaks that context.
How to design a quiz that earns its place.
Decide formative or summative first.
A formative quiz reinforces learning — low stakes, immediate feedback, no pass mark. A summative quiz measures learning — higher stakes, feedback withheld during, pass mark applied at the end. Don't mix the two. A quiz that pretends to be casual then enforces a pass mark feels like a trick.
Three to ten questions is the sweet spot.
Fewer than three and you can't measure anything reliably — a single wrong answer becomes a third of the score. More than ten and learners fatigue, especially in compliance settings where the quiz follows a long module. If you need more questions for psychometric reasons, split them into multiple shorter quizzes spaced across the course.
Mix question types deliberately.
All single-answer multiple choice gets monotonous. A mix of single-answer, multi-select, true/false, and the occasional drag-and-drop or hotspot question keeps engagement. But each question type should be chosen because it's right for what's being tested — see Writing good assessment questions for which format fits which content.
Per-question feedback that teaches.
"Correct!" and "Incorrect!" are wasted moments. Per-question feedback should explain why an answer is right — and ideally, why each distractor is wrong. This turns assessment into teaching. If the quiz is summative and you're saving feedback for the end, write the end-of-quiz feedback so it goes question by question, not just a score.
Decide whether learners can review their answers.
For formative quizzes — yes. Let learners scroll back through their answers and feedback before continuing. For summative quizzes — no. Review during the quiz lets learners change strategy mid-way; review after the pass mark is calculated is fine.
Random order — but only when it helps.
Randomising question order across attempts stops learners memorising "Question 3 is C." Randomising option order within a question stops them memorising "the answer is always option B." Useful for re-attemptable quizzes. Not useful — and sometimes harmful — when questions have an intentional sequence or when options have a natural order (severity, time, size).
Handle "didn't pass" with care.
The screen after a failed attempt is the most important screen in the quiz. Be kind. Tell the learner what they missed, give them a path back to the relevant content, and let them try again without making it feel like punishment. A failed quiz is an opportunity to teach — taking that opportunity well builds learner trust.
When you'd reach for a quiz.
Mid-module knowledge checks
Three to five questions after a content section, low stakes, immediate feedback. Retrieval practice — one of the strongest learning interventions available.
Compliance certification
Final summative assessment with a defined pass mark (usually 80%). Pass = certificate, fail = retake. Be deliberate about whether the learner sees their wrong answers before retaking.
Pre-course self-assessment
Three to five questions at the start of an optional course, no pass mark, used only to recommend "start here" or "skip ahead". Never graded against the learner.
Spaced retrieval over time
The same short quiz delivered three times: at the end of the course, two weeks later, and two months later. Spaced retrieval is one of the most evidence-backed learning interventions there is.
Want a quiz
that actually measures?
We design assessments that test understanding, not test-wisdom. Three-week delivery, transparent pricing, zero drama.