Start any writing task with a seven-minute keep, cut, question strip. Ask students to share one cut and why it does not serve the purpose.
This forces judgement before any automation happens. Students who cannot complete the triage cannot have genuinely read the materials. That is the integrity signal.
When AI synthesis is driven by poor context, the output reflects the context, not the task. A student who pastes everything available gets confident-sounding output that misses the point. A student who triages first gets output that is actually fit for purpose.
Teaching context triage is teaching the prior skill that makes AI a useful thinking partner rather than a confident generator of irrelevance.
After reading a short text set, students list three keep facts, two cut details, and one question claim that needs checking. Only then do they write, discuss, or generate options.
Students submit a one-page context triage alongside their draft: which sources were used, which were rejected, and why the final set is fit for purpose. Makes thinking visible without surveillance.
Are students using only class-approved, locally grounded evidence, and do they understand what must be excluded for privacy and cultural safety?
Context triage creates a natural moment to teach two questions: what should never go in, and what perspectives are absent from this evidence set? Both are design questions, not compliance questions.
Print or save this resource as a PDF using your browser.