0
Analogy generator stress-tested for where it breaks down
⁂auto-checked, 3 hours oldAauraNovice
The prompt
Generate an analogy to explain concept to audience, but don't just give me the analogy, stress-test it so I know its limits before I use it in front of students.
What aspect of the concept is hardest for this audience to grasp (the actual sticking point, not the whole topic): specific_sticking_point
Do this:
1. Propose an analogy that specifically targets the sticking point, not a generic analogy for the whole topic.
2. Walk through how the analogy maps onto the real concept, point by point, so I can see exactly which feature of the analogy represents which feature of the real thing.
3. Now find where the analogy breaks down, every analogy does eventually. Identify the SPECIFIC point past which a student might draw a false conclusion if they extend the analogy too far, and state what wrong idea they'd end up with.
4. Tell me whether I should proactively name that breaking point when I introduce the analogy, or whether it's fine to let it go unstated because students are unlikely to push it that far unprompted, use your judgment on this and explain why.
5. If you can, offer a second, different analogy for the same sticking point, sometimes offering two analogies that break down in different places gives students a more accurate composite picture than either alone.
Where it has been run
| Model | Version | People | Broke | Auto | Last |
|---|---|---|---|---|---|
| Claude | Haiku 4.5 | 0 | 0 | 1/1 | 3 hours ago |
Example output
Sticking point: students confuse correlation and causation. Analogy proposed: ice cream sales and drowning deaths both rise in summer, not because one causes the other, but because heat drives both. Mapping: ice cream sales and drownings represent the two correlated variables, summer heat represents the hidden common cause (confound), the analogy directly targets the sticking point rather than explaining correlation generally. Breakdown point identified: a student might extend this to assume every correlation always has a hidden common-cause explanation, which is false, sometimes one variable genuinely does cause the other directly, and sometimes correlation really is coincidental with no causal link at all. Recommendation: proactively name this limit when introducing the analogy, since 'there's always a hidden third cause' is a specific, plausible wrong conclusion students commonly draw from exactly this kind of example, worth heading off explicitly rather than leaving implicit. Second analogy offered: a rooster crowing before sunrise, useful because it demonstrates the specific case of coincidental sequencing without any causal link at all, complementing the first analogy's hidden-cause case.
0 comments
Sign in to comment or report what this prompt did for you.
Sign inNo comments yet. Run the prompt and report what you got.