Most assessment questions in corporate eLearning measure one thing: whether the learner can remember what was on the previous screen. That’s recall — not learning. And it produces a dangerous illusion: high assessment scores that mask the fact that nobody can actually apply what they were taught.
The difference between a useful assessment and a useless one isn’t difficulty. It’s alignment. A good question tests whether the learner can do the thing the training was designed to teach them to do. Here’s how to write assessments that measure real learning — and give you data you can actually use.
The problem with recall questions
A recall question looks like this: “According to the company’s data privacy policy, personal information must be stored for a maximum of how many years?” The learner either memorized the number or they didn’t. If they memorized it, they pass. If the number was on the screen three minutes ago, most people pass. None of this tells you whether the learner will actually handle personal data correctly when they encounter it in their work.
Recall questions test short-term memory, not competence. They reward learners who pay attention to details on screen and penalize learners who focus on understanding concepts. Worst of all, they give organizations a false sense of confidence — 90% pass rates that mean almost nothing for on-the-job performance.
Write questions at the application level
The fix is writing questions that require application — not recall. An application-level question presents a realistic situation and asks the learner to make a judgment or take an action based on what they learned.
Instead of “what is the maximum storage period for personal data?” try: “A client requests that you retain their personal records indefinitely for their convenience. Based on the data privacy policy, what should you do?” Now the learner has to apply the policy to a realistic situation — which is exactly what they’ll need to do on the job.
The difference is context. Recall questions strip away context and test isolated facts. Application questions embed facts in situations where the learner has to use them. One measures memory. The other measures judgment.
Use scenario-based questions for complex skills
For training that involves decision-making, problem-solving, or interpersonal skills, scenario-based assessment is the gold standard. A scenario presents a realistic situation with enough detail to require genuine thinking, then asks the learner to choose an action or sequence of actions.
A good scenario has three elements. First, a realistic setup that mirrors what the learner encounters in their actual role — not a textbook example, but something they’d recognize from their last workweek. Second, plausible options where the wrong answers aren’t obviously wrong. If the learner can eliminate three options without thinking, the question isn’t testing judgment — it’s testing elimination. Third, meaningful feedback that explains why each option would lead to a specific outcome, not just “correct” or “incorrect.”
Scenario questions take longer to write than recall questions. They’re worth it. A five-question scenario assessment tells you more about a learner’s readiness than a twenty-question recall quiz.
Match the question format to the objective
Different learning objectives require different assessment formats. Using the wrong format produces unreliable results.
For factual knowledge — definitions, procedures, regulations — multiple choice works if the questions are written at the application level. Present the fact in context and ask the learner to apply it, not just identify it.
For procedural skills — completing a workflow, following a process, using a system — use sequencing or ordering questions. Ask the learner to arrange steps in the correct order, or identify which step comes next in a given scenario. Better yet, use a simulation where the learner performs the procedure.
For judgment and decision-making — handling exceptions, managing difficult situations, applying guidelines to ambiguous cases — use branching scenarios where each decision leads to a different outcome. The assessment isn’t a single question but a series of connected decisions that reveal how the learner thinks.
For interpersonal skills — coaching conversations, de-escalation, feedback delivery — use video or audio scenarios where the learner observes a situation and chooses how to respond. Written descriptions of interpersonal situations lose the nuance that makes these skills difficult in real life.
Write wrong answers that teach
In most corporate eLearning, wrong answer options are obviously wrong — sometimes absurdly so. “What should you do if a customer complains?” with options including “ignore them” or “argue with them” isn’t assessing anything. The learner doesn’t need training to eliminate those options.
Effective distractors — the wrong answer options — represent common mistakes, misconceptions, or partially correct approaches that a learner might genuinely choose if they haven’t fully understood the material.
If the training teaches a four-step de-escalation process, the wrong answers shouldn’t be “yell at the customer.” They should be approaches that skip a step, apply the steps out of order, or address the wrong element of the situation. When a learner selects one of these options, the feedback can explain exactly where their thinking went wrong — which turns the assessment into a learning moment.
Use pre-assessments to personalize the path
Pre-assessments given before the training starts serve two purposes. They establish a baseline for measuring improvement — you can compare pre and post scores to quantify learning gain. And they identify what the learner already knows, which allows you to skip content they don’t need and focus on genuine gaps.
A well-designed pre-assessment uses the same question types and difficulty level as the post-assessment. If the learner passes a section of the pre-assessment, they can skip that section of the training. This respects the learner’s time, reduces unnecessary content consumption, and focuses the training on the areas where it’s actually needed.
Measure what matters — then act on it
Assessment data is only valuable if someone analyzes it and acts on the findings. After every training deployment, review the assessment results for patterns. Which questions have the highest failure rates? Those topics need better content or more practice. Which questions does everyone pass easily? Those might be too easy, or the content might be covering something the audience already knows.
The assessment isn’t just a gate the learner passes through. It’s a diagnostic tool that tells you whether your training is working — and where it needs to improve.
