Methods · Generative practice & retrieval
B19
Distractor analysis
Say why each wrong option is tempting, not only why the right one is right, so multiple choice stops being a recognition game.
When to use it
- You get MCQs right by vibe and cannot explain the traps.
- Wrong options feel “also kind of true” and you want that named.
- Past papers punish near-miss misconceptions the course deliberately plants.
- You can justify the key but not inoculate yourself against the distractors.
What to ask AI
Type something like: “Here’s an MCQ. I’ll pick an answer and explain why each wrong option is tempting. Challenge my distractor reasons before you confirm the key.”
Ask it not to reveal the correct option until you have reasoned about the traps.
For each distractor, demand a misconception name in plain language.
How it helps
Distractors are wrong answers designed to look right if you hold a common misconception.
Knowing why the key is right is only half the exam skill.
Explaining each trap builds discrimination: you see the exact wrong belief the option feeds.
That belief is often the same one that will trip you in short-answer form later.
How to run it
- Attempt the question and commit to an option.
- For every other option, write why a student might pick it.
- Ask the AI to pressure-test those reasons.
- Only then confirm the key and refine the misconception names.
- Add recurring misconceptions to your error list and remediation.
Example exchange
I pasted an MCQ. I pick B. My read of the distractors: A tempts people who confuse precision with accuracy; C tempts unit errors; D tempts reversing a definition. Challenge my distractor stories before you confirm the key.
Your story for A is clear. For C only: which unit conversion exactly would make someone pick that option?
Mixing metres per second with kilometres per hour in the stem numbers.
That holds. Now one line for D: quote the reversed definition that would make D look right.
Treating “necessary” as if it meant “sufficient.”
Solid. The key is B. The costly trap in this paper’s style is D, schedule a one-item remediation on necessary versus sufficient.
Copyable prompt
Here is an MCQ:
[PASTE]
I will choose an answer and explain why EACH wrong option is tempting (name the
misconception). Do not reveal the key until I finish. Then challenge weak
distractor reasons.
The Tell
You will know the method has flipped when you only justify the right option, or you let the AI narrate the distractors while you nod along.
Then you practised recognition of a good explanation again.
Write a temptation story for every wrong option before any confirmation.
Principle evidence
Strength of the underlying learning idea, not a claim about AI products.
Analysing attractive wrong answers supports conceptual discrimination and error awareness. It aligns with research on comparing correct and incorrect examples and on misconception-focused feedback.
AI delivery evidence
Whether an AI tutor delivers this method well is a separate question.
AI may invent clever distractor stories that were not the item-writer’s intent. Prefer your course’s common errors; treat AI colour as optional.