Verified against ChatGPT · 2026-08-10
Write MCQs where every wrong option represents a real misconception, not filler
Generates multiple-choice practice questions with distractors built from actual common errors on the topic, so getting one wrong tells you which specific misunderstanding you have instead of just that you guessed wrong.
The prompt
Ready to copy — highlighted parts are example details you can swap.
Write a set of multiple-choice questions on the topic below where every incorrect option corresponds to a specific, real misconception someone could plausibly hold — not a random wrong number or an obviously silly option included just to fill four slots. TOPIC Balancing redox equations using the half-reaction method EXAM CONTEXT NEET UG Chemistry, single-best-answer format, roughly 1 minute per question under exam conditions KNOWN COMMON MISCONCEPTIONS ON THIS TOPIC Forgetting to balance oxygen atoms using water before balancing hydrogen with H+ in acidic medium; adding electrons to the wrong side of a half-reaction NUMBER OF QUESTIONS 8 CONSTRUCTION RULES Each question must have exactly four options with exactly one correct answer. For every distractor, it must be traceable to a specific, nameable error — a sign error, a units mistake, an off-by-one in a formula, confusing two similar-sounding terms, applying the right method to the wrong step, or a misconception from the list I gave you. Never use "all of the above," "none of the above," or a distractor that's simply implausible on its face (e.g. wildly out-of-range numerically) — every option should require the test-taker to actually check their work to rule it out. Vary which position (A, B, C, or D) holds the correct answer across the set so the answer key isn't visually patterned. Do not tell me which option is correct anywhere near the question itself. WHAT NOT TO DO Do not write a question stem that gives away the answer through phrasing (e.g. a stem that only grammatically fits one option). Do not repeat the same misconception as the basis for a distractor more than twice across the set — vary which error each question is probing for. OUTPUT FORMAT Section 1: numbered questions, each with four lettered options, no answer indicated. Section 2, headed ANSWER KEY: for each question, the correct letter plus, for each of the three wrong options, a one-line note naming the specific misconception that option represents.
Customize
Optional — swap in your own details for the highlighted parts above.
Why this works
Left unguided, GPT-5.1's default distractor-writing habit is to generate one plausible near-miss and then pad the remaining two or three options with numerically or logically implausible filler, because a genuinely misconception-grounded distractor requires modeling a wrong mental process rather than just perturbing the right answer slightly — the explicit instruction to trace every option to a specific nameable error (sign flip, unit confusion, wrong formula step) forces the harder generative task instead of letting it default to the easier one. This matters diagnostically, not just for difficulty: an MCQ set built from real misconceptions turns a wrong answer into information about which specific error you're making, while an MCQ set with implausible filler distractors only tells you that you got it wrong, with no diagnostic value at all — which defeats the actual purpose of practicing with MCQs rather than open-response questions. Banning "all/none of the above" addresses a separate, well-documented pattern where these options let a test-taker (and the model) avoid fully committing to four independently-justified wrong answers, effectively reducing the question to three real options plus an escape hatch. The instruction to vary correct-answer position and avoid repeating the same misconception basis more than twice guards against a subtler failure mode: GPT-5.1 tends toward repetitive structural patterns across a batch of similar items unless explicitly told to vary them, which would otherwise let a test-taker pattern-match the answer key rather than the content.
What you get back
Q4: In acidic medium, which is the correctly balanced half-reaction for the reduction of MnO4- to Mn2+? ... Answer key Q4: Correct = C. Option A represents forgetting to balance oxygen atoms with water molecules before adding H+. Option B represents adding electrons to the reactant side instead of the product side. Option D represents a charge-balance error from miscounting the electrons needed.
Verified against
ChatGPT GPT-5.1 · 2026-08-10
Changelog
- 2026-08-10 — Initial publish, verified against ChatGPT GPT-5.1.
Need this built into your business?
If a prompt isn't enough — what Scult builds, built and maintained for you — that's Scult's day job.
EXPLORE WHAT SCULT BUILDS
