Most training assessments are easier to pass than they should be. A participant who skimmed the slides, remembers a few keywords, and has a knack for spotting the most official-sounding answer can often score well without understanding much at all.
That matters, because the point of an assessment is not to produce a score. It is to tell you, and the learner, whether they can actually do what the training was meant to teach. Here is how to write questions that test understanding rather than recall.
Start from what people need to do
Before writing a single question, write down what people should be able to do after the training. Use action verbs: diagnose, choose, prioritize, explain, apply. Avoid "know" and "understand," which are impossible to observe directly.
For example, a customer service objective might shift from "Understand the refund policy" to "Decide whether a customer's request qualifies for a refund and explain why." The second version immediately suggests better questions.
Use scenarios instead of definitions
Questions that ask "What is X?" test memory. Questions that ask "Here is a situation, what should you do?" test understanding.
Compare these two:
- Memory: What is the maximum number of days for a standard return?
- Understanding: A customer bought a jacket five weeks ago, wore it once, and wants to return it because the zipper broke. What do you offer, and why?
The second question requires people to recall the policy, interpret it, and apply it to a messy, realistic situation. That is what the job actually requires.
Write better wrong answers
In multiple choice questions, the wrong answers (often called distractors) do most of the work. Weak distractors are obviously wrong, which makes the right answer easy to spot. Strong distractors reflect real misconceptions.
To write good distractors:
- Use mistakes you have actually seen people make on the job.
- Make every option plausible and similar in length and tone.
- Avoid "all of the above" and "none of the above," which reward test-taking strategy.
- Do not make the correct answer the longest or most detailed option. People notice.
Tip: If you ran a pre-quiz or a live session, look at the wrong answers people chose most often. Those are your best distractors for the final assessment.
Go beyond multiple choice
Multiple choice is efficient, but it is not the only option. Different question types reveal different kinds of understanding:
- Sorting questions show whether people know the correct sequence of a process, not just its parts.
- Matching questions test whether people can connect concepts to examples, causes to effects, or problems to solutions.
- Find on image and label an image questions test whether people can recognize something in context, such as a fault on equipment or a field on a form.
- Open questions ask people to explain their reasoning in their own words. They take longer to review but reveal a lot.
- Prioritization questions test judgment by asking people to rank actions by urgency or importance.
Ask "why," not just "what"
One simple upgrade: after a scenario question, add an open follow-up asking people to explain their choice. You will quickly spot people who chose correctly for the wrong reasons, and people who chose wrongly for reasons that point to a gap in the training.
Check your questions before you use them
A few final checks:
- Could someone answer this correctly without taking the training? If yes, it may be testing general knowledge or common sense.
- Could someone who understood the training get it wrong because of confusing wording? If yes, rewrite it.
- Does every question map to an objective? If not, cut it.
- Is there a good mix of question types and difficulty levels?
Ask a colleague who knows the subject to take the assessment before you roll it out. Their questions about wording will save you trouble later.
Use results to improve the training
Assessment results are feedback for you as much as for learners. If most people miss the same question, either the question is flawed or the training did not cover that point well. Both are worth fixing.
Building assessments in LetMeAdd
LetMeAdd gives you a wide range of question types for building assessments that test real understanding:
- Combine multiple choice, sorting, matching, fill in the blanks, find on image, label an image, prioritization, and open questions in one assessment.
- Run it live with participants joining by 6-character code or QR code, no account needed, or share it as a self-paced quiz people complete from a link, with scores.
- Start from the ready-made training assessment template and adapt it to your objectives.
- On the Pro plan, use AI question generation to draft questions, then review and edit them before adding.
- Export results per session to analyze which questions people struggled with.
Learn more on our online quiz maker page.
An assessment that tests understanding takes a little more effort to write. It is worth it, because it finally tells you whether your training worked.
Build an assessment that measures what matters. Create your LetMeAdd account.