Multiple-choice question review: three worked examples https://ai-essay.com/guides/multiple-choice-question-checklist Original fictional practice material. S1 and S2, the draft questions, proposed keys and repairs below were constructed for this resource. They are not a real campus notice, captured AI output, student responses or a classroom test. S1 — The campus library offers evening workshops, online guides and bookable appointments. All three are free for enrolled students. This notice reports no attendance figures or learning outcomes. S2 — A workshop consists of two parts lasting 20 minutes each, with one 10-minute break between them. Example 1: a draft with two defensible answers Objective: Identify the stated fee policy for the listed library services. Source: S1 Draft: Which support is free for enrolled students? A — Evening workshops. B — Bookable appointments. C — Paid external coaching. D — Private commercial tutoring. Draft key: A Your diagnosis and repair before reading on: Possible editorial review: Revise before use A and B are both supported by S1, so the draft key cannot make this a single-best-answer item. Adding the instruction choose the best answer would not resolve the missing distinction. Repaired item: According to S1, what is the fee policy for the three listed services? A — All three are free for enrolled students. B — Only workshops are free for enrolled students. C — Only guides are free for enrolled students. D — Each service has a fee for enrolled students. Repaired key: A A preserves the bounded policy for the three services and the enrolled-student condition. B narrows the free offer to workshops, contradicting the stated coverage of all three. C narrows the free offer to guides, contradicting the same coverage. D adds a fee that the notice explicitly rules out for enrolled students. Status: constructed repair for inspection; no independent teacher review, learner trial or release has occurred. --- Example 2: a key that the source cannot establish Objective: Distinguish information supplied by a notice from information it leaves unreported. Source: S1 Draft: What percentage of enrolled students attended the workshops? A — 20%. B — 40%. C — 60%. D — 80%. Draft key: B Your diagnosis and repair before reading on: Possible editorial review: Hold the unsupported claim; replace the item S1 gives neither attendance counts nor a participation rate. No listed percentage can be justified. A plausible-looking number is not a recoverable answer; changing which percentage is marked would leave the same problem. Repaired item: Which information is missing from S1? A — The available support formats. B — The cost for enrolled students. C — The number of workshop attendees. D — The option to book an appointment. Repaired key: C A is supplied by the list of workshops, guides and appointments. B is supplied by the free-access statement for enrolled students. C is absent: the notice expressly reports no attendance figures. D is supplied by the phrase bookable appointments. Status: constructed repair for inspection; no independent teacher review, learner trial or release has occurred. --- Example 3: two options express the same duration Objective: Calculate total elapsed duration, including the stated break. Source: S2 Draft: How long is the entire workshop in S2, including the break? A — 40 minutes. B — 50 minutes. C — 60 minutes. D — 5/6 hour. Draft key: B Your diagnosis and repair before reading on: Possible editorial review: Revise the option set before use The total is 20 + 10 + 20 = 50 minutes. B and D are equivalent because (5/6) × 60 = 50. The draft therefore has two correct choices despite a single marked key. Repaired item: How long is the entire workshop in S2, including the break? A — 20 minutes. B — 40 minutes. C — 50 minutes. D — 60 minutes. Repaired key: C A counts only one of the two workshop parts. B counts both parts but leaves out the break. C includes two 20-minute parts and the single 10-minute break: 50 minutes. D would add a second 10-minute break that S2 does not specify. These are possible error paths for reviewing the distractors, not observed learner diagnoses. Status: constructed repair for inspection; no independent teacher review, learner trial or release has occurred. Method reading (supports the review method, not the fictional notices) https://uwaterloo.ca/centre-for-teaching-excellence/catalogs/tip-sheets/designing-multiple-choice-questions https://www.cmu.edu/teaching/assessment/assesslearning/creatingexams.html https://crlt.umich.edu/designing-effective-exams https://teaching.uic.edu/cate-teaching-guides/digital-learning/adapting-assignments-for-the-age-of-ai/ https://www.cmu.edu/teaching/assessment/assesslearning/concepTests.html