Purpose: One narrative scenario per high-risk assumption. Specific enough that a program sponsor reading it would say 'that would be a serious problem' — not 'I suppose that's possible.' Vague scenarios are not disconfirmation scenarios.
Disconfirmation scenarios are narratives, not risk statements. "Managers do not complete the observation protocol" is a risk statement. A disconfirmation scenario shows: who found out, when, what data made it undeniable, and what became impossible as a result. The specificity activates a different cognitive response — and reveals design changes that forward-looking analysis misses.
Scenario Format Guide
For each scenario, answer these five questions:
1. It is [N] months after launch — what specifically happened? | 2. Who discovered it, and in what context? | 3. What data made it undeniable? | 4. What was the organizational consequence? | 5. What became impossible without this assumption being true?
Then ask: "If we knew this scenario would happen, what would we design differently?"