WATER Check · Teacher guide · Sept. 14, 2026 · Version 1.0
Four complaint forms. Twice the story.
A free journalism and AI literacy lesson for grades 6–12. Students investigate a fictional lunch-line claim, separate counts from rates, trace repeated claims, and defend an editorial decision.
Before class
Time: 45 minutes, plus an optional 15-minute transfer task. Materials: Student handout, paper, and a calculator if helpful. The printed source packet is complete; internet access, AI accounts, and the book are optional.
Learning evidence: A student can distinguish a count from a rate, show where a claim comes from, explain the limits of the supplied evidence, and write an attributed response proportionate to that evidence.
Print or distribute the student handout without this answer key. All cases are visibly labeled simulations. Tell students that “not established” is a legitimate finding when they explain the missing evidence. The same rate does not prove the school has no lunch problem.
45-minute sequence
| Time | Teacher move | Student evidence |
|---|---|---|
| 0–5 minutes | Display Item 1. Ask students to record a private reaction, then separate its claims. Do not take a public vote on who was fooled. | Initial reaction and three distinct claims. |
| 5–12 | Introduce Wait, Ask, Trace, Examine, Respond. Give students all four source items. Model the difference between “the count doubled” and “the schedule caused harm.” | Evidence needed for each claim. |
| 12–22 | Pairs trace the repost’s links and calculate forms per lunch. Ask what the denominator measures. | A source map and two calculations with units. |
| 22–32 | Release the generated-style summary. Remind students it is instructor-written. Ask them to audit every claim against the packet. | Four claim judgments and a repaired summary. |
| 32–40 | Assign reporter and editor roles. The editor asks: “What can we publish, and what still needs reporting?” Then switch roles. | A two-sentence attributed brief and an editorial decision. |
| 40–45 | Collect an individual exit note. Ask which missing evidence would most change the conclusion. | Independent explanation of a revision and next step. |
Teacher key: Lunch line
The count of complaint forms doubled, from 4 to 8. The rate of forms per lunch did not: 4 ÷ 800 = 0.005 = 0.5%; 8 ÷ 1,600 = 0.005 = 0.5%. These are forms per lunch, not the percentage of students who complained or were dissatisfied.
Item 4 points back to Items 1 and 2. That link structure does not establish independent confirmation of the broad conclusion. The packet does not disclose whether Item 1 used separate reporting. Students should not invent that history.
The records do not establish waiting times, every student’s experience, causation, or whether reversing the schedule would solve a problem. Unchanged forms per lunch also cannot establish unchanged service quality. The reporting system or who submits a form might have changed.
Summary audit: “Doubled the complaint rate” is contradicted for forms per lunch using the supplied records. “All students” is not established. “Two independent reports” is not established. “Reversing the schedule will solve the problem” is not established.
Possible brief: “The school bulletin reports 8 lunch complaint forms in October, up from 4 in September. With the office tally showing twice as many lunches served, forms per lunch remained at 0.5%; these records do not establish changes in wait times or the schedule’s effect on service.”
A defensible next step is to seek comparable wait-time observations, check whether complaint procedures changed, and interview students with different experiences. Accept different publication decisions when the evidence and limits are explicit. This is a reasoning exercise, not a contest to match the teacher’s headline.
Assessment: 0–8 points
Score each dimension 0, 1, or 2. This is a proposed classroom rubric, not a validated measurement instrument.
| Dimension | 2 points | 1 point | 0 points |
|---|---|---|---|
| Claim | Separates count, experience, and causal claims. | Identifies a checkable claim but blends interpretations. | Repeats the conclusion without a checkable claim. |
| Trace | Maps supplied links and explains the limits of independence. | Names records without explaining dependencies. | Treats repetition as independent confirmation. |
| Context | Calculates both rates with correct units and names a material limitation. | Finds the numerical issue but misses units or limitations. | Makes an unsupported count-to-rate or causal inference. |
| Response | Gives an attributed, qualified response and a useful next reporting step. | Corrects part of the claim but leaves attribution or next steps unclear. | Publishes an unsupported conclusion or dismisses all evidence. |
Independent practice: Library attendance
Use the separate library handout after teaching the lunch case, or at the next lesson. Complete it individually before discussion. Visits tripled, but visits per session stayed at 2: 6 ÷ 3 and 18 ÷ 9. Visits are not unique visitors or reading time. Causation, universal effects, independent corroboration, and guaranteed future gains are not established.
Compare the same rubric dimensions and record where students transfer the reasoning. The cases have similar structures; they have not been shown to have equal difficulty. A higher practice score alone does not establish that WATER caused improvement.
Adaptations and extensions
- Middle school: Supply a two-row calculation table, define “rate” and “denominator,” and permit a spoken explanation or paired scribe.
- High school journalism: Require a source note, a headline, an editor’s challenge, and a correction that reaches the same audience as the original error.
- English and civics: Annotate the phrases that turn numbers into conclusions. Rewrite the paragraph while preserving the supported facts.
- Libraries: Use the library case with a community group. Ask what a program report needs to explain its attendance figures fairly.
- AI extension: Audit a teacher-selected, saved answer under local policy. Record its date, prompt, and system if known. Verify factual claims against sources, not the answer’s confidence. No student account is necessary.
Next steps
Use the existing DHMO demonstration to compare the same problem in a different setting. Try the guided WATER Check on a teacher-selected claim. On a shared device, copy or print the work and clear it before leaving.
Explore the developing semester courses or ask about a workshop. To share feedback, use the WATER feedback form; leave out student names and identifying details.
Source and status
This lesson adapts Cases A and B and the assessment approach from the existing News Media and AI Literacy course design dated Sept. 6, 2026. Framework: Eric Jourgensen, WATER v2.0. Lesson packaging and adaptations prepared with AI assistance. The simulated summary is not evidence about any actual AI system. This resource does not establish research validation or institutional endorsement. Standards alignment is not claimed for this standalone lesson.
Use your browser’s Print command for a paper copy or PDF. The student pages print separately without the key. Return to teaching resources.