← Credibility and Verification
Tomorrow Ready · Credibility and Verification

When AI Agrees Too Easily: Verification Slip for Social Sciences, Years 4 to 6

Subject adaptation · Years 4–6 · Social Sciences · Field-Based STEM · Tony Jones

AI systems produce answers that sound authoritative even when they are wrong, and they tend to agree with the position the user has already taken. A Verification Slip requires students to check one claim against an independent source before the work is complete.
1Name
One Claim
2Find
One Source
3State
One Limit
4Attach
to Work
The Strategy
  1. Introduce the Verification Slip as a half-page document that travels with any social inquiry writing.
  2. Students name one specific claim their work makes: a factual assertion, a cause-and-effect statement, or a claim about who does or thinks something.
  3. Students find one class-approved source that supports that specific claim, naming the title and where they found it.
  4. Students state one limit: a condition under which the claim might not hold, a group it might not apply to, or something the source does not address.
  5. The slip is submitted with the work and assessed as a component of it.
Year-Band Practice

Years 4–5

Teacher selects two or three class-approved sources in advance. The student matches their claim to the right source. The limit is guided: "One thing this source doesn't tell us is..." scaffolds the thinking without removing it.

Year 6

Students identify their own source from a class-approved list. The limit sentence is written without scaffolding. This builds independence ahead of Years 7 to 8 inquiry demands.

AI Use Note

AI systems can produce plausible-sounding claims that agree with the user's existing position. Students using AI for social science research may be reading agreement rather than evidence. The slip catches this by requiring an independent check.

Implementation

Decision Checkpoint

Run the Verification Slip for any writing task where students have used AI to research or gather information. AI systems can produce plausible social science claims about Aotearoa New Zealand communities that are factually incorrect or significantly simplified.

Teacher Judgement Note

When a student cannot find an independent source for their claim, treat this as a teaching moment about claim quality. Some AI-generated claims cannot be verified because they do not correspond to any real source.

Related Frameworks

Verification Slip · Context Triage · Automation Is Not Research

Tony Jones · Founder, Field-Based STEM · Tomorrow Ready Resources · Free to use and share