Exam design

How to write multiple choice questions that reveal learning

A multiple choice question presents a problem—the stem—and a set of answer options. One or more options are keyed as correct, while distractors represent plausible but incorrect responses. Strong items can assess more than recall when students must interpret evidence, apply a rule, or choose among defensible decisions.

Anatomy of a multiple choice item

PartPurposeDesign test
StemStates the problem or questionCan a prepared student understand the task before reading the options?
KeyThe best or correct responseIs it defensible without relying on hidden assumptions?
DistractorsRepresent plausible errors or alternativesWould each distractor attract a student with a specific misconception?
ContextProvides data, scenario, text, image, or constraintsIs every detail needed to answer the question?

Write a strong item

  1. Start with the learning outcome and identify the decision or reasoning you want to observe.
  2. Write the correct response before the distractors. Verify that it remains correct under the stated conditions.
  3. Build distractors from common errors, incomplete reasoning, or nearby concepts—not random wrong answers.
  4. Keep options parallel in grammar, specificity, and approximate length.
  5. Remove clues such as absolute wording, repeated phrases, unequal detail, or a key that is consistently longest.
  6. Ask a colleague or student assistant to explain how each option might be chosen.

Higher education item patterns

  • Interpretation: provide a graph, excerpt, dataset, code sample, or clinical finding and ask for the best interpretation.
  • Application: describe a new case and ask which principle, method, or next action applies.
  • Error analysis: show a worked solution and ask where the reasoning first becomes invalid.
  • Comparison: present two approaches and ask which better meets stated constraints.
  • Prediction: ask what is most likely to happen when one condition changes.
  • Best justification: provide a decision and ask which evidence most strongly supports it.

Common mistakes

  • Testing trivia that is easy to write but unrelated to important outcomes.
  • Using “all of the above,” which can reward partial recognition rather than full reasoning.
  • Writing negative stems without a genuine reason; if necessary, emphasize the negative term.
  • Creating trick questions whose difficulty comes from wording rather than the subject.
  • Using overlapping options or more than one defensible key.
  • Reusing exposed questions without reviewing whether they still provide valid evidence.

After administration, review both statistics and student reasoning. A difficult item is not automatically good, and an easy item is not automatically weak. Look for unexpected option patterns, ambiguity, content that was not taught, and distractors no one selected. Instructor review remains essential before changing a grade or reusing the item.