Teacher Guidelines
Overview
| Element | Implementation | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Time | 2–3 lessons + homework | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Quality band | B (practice) or C (full project with submission) | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Group size | Individual or pair work | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Tools | AI tool with web search; access to journals/retraction databases available to the school | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| (Fill-in template table) | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Why This Theme Works
It teaches one of the most important skills of AI literacy: tracing a claim to its source and evaluating its status, instead of trusting confident phrasing. It suits science-curious students and offers real depth for those who finish quickly (see Aron's user story).
Preparation
- Ensure access to an AI tool with web search.
- Show students how to check the status of a source: search for the article, check if it is retracted (official retraction database / the journal itself), and whether the result has been replicated. If the school has source tools (e.g., Scite), they can be used.
- No tool required that the student needs personal access to — a free account or open tools are enough.
- Read the completed example for the expected quality standard.
Progression
- Step 1 (15 min, without machine): students write a fact + how sure they are.
- Step 2 (lesson): the ICCRA prompt; trace to source, get sources with status.
- Step 3 (lesson + homework): verify the status of each source, judgment.
- Discussion: who found a hallucinated source? Who found a fact that rested on nothing?
Common Pitfalls
- Source considered proof without status being checked. A retracted article still "exists" — but it no longer supports the claim.
- Hallucinated citations. Machines invent authors, years, and DOIs that look convincing. Require a lookup of every source.
- Consensus confused with echo. Multiple sites repeating the same study is one source.
- "I don't know" as a failure. On the contrary — being unable to verify is a valid, honest result and should be recognized.
Adaptation
- Natural Sciences: delve into the replication crisis and peer review as a process.
- Vocational studies: choose a fact from the trade (e.g., about materials, safety, or methods) and trace it.
- Shorter version: one fact, one source of origin, confirmed or refuted.
Discussion Questions
- What is the difference between "many people say this" and "this is proven"?
- Why can a machine be so confident about something that does not hold up?
- Is it more honest to say "I don't know" than to repeat a claim with confidence?