A PIO drafts a boil-water notice, prompts a critique for panic language, then rewrites a calmer version. What technique is illustrated?

Public alert wording improves through critique-and-revise cycles. Drafting a boil-water notice, prompting a critique for panic language, then rewriting a calmer version illustrates iterative refinement: generate, evaluate against a risk (panic wording), then revise. For a PIO, that loop protects clarity and actionability—what to do, what water to use—without amplifying fear that drives rumor or overload of emergency lines.

Sampling multiple parallel drafts with best-of-n selection for public alert wording fails because the scenario relies on critiquing and revising a single draft, not generating multiple parallel drafts to select the best from. Assigning a resident persona with community voice prompts for public alert wording fails because the scenario describes a PIO revising official alert wording through critique, not adopting a resident persona or shifting to community voice. Adding step-by-step reasoning chains to refine tone for public alert wording fails because the scenario focuses on revising alert tone through critique-and-revise prompts, not on generating explicit step-by-step reasoning chains.

Exam caveat: calmer wording is not the same as understating hazard. Iterative refinement should remove panic rhetoric while preserving mandatory risk facts, timelines, and protective actions required by public-health protocol.

Operational check: run a panic-language critique checklist (absolutist scare phrases, missing action steps, conflicting instructions); revise until facts and actions remain and panic flourishes are gone; obtain PIO or incident-commander approval before release.