The business resolution metric adequately represents: Which prior assumption no longer holds?
A business metric can remain strong while affected employees experience poor usability or harm, so the metric’s sufficiency assumption fails.
The question
An internal employee-assistance chatbot reports a high resolution rate. After expansion to employees with limited digital literacy, users report confusing and occasionally harmful guidance despite the business metric remaining strong. Which prior assumption no longer holds?
Preparing for AIGP? Take the free 5-min readiness quiz →
- The business resolution metric adequately represents employee outcomes. ✓Strong resolution statistics no longer capture usability and harm concerns among the expanded population, requiring broader outcome assessment.
- Resolution rate is unrelated to employee experience.Resolution rate may correlate with experience, but the scenario does not establish that it is wholly unrelated or irrelevant.
- Digital-literacy differences affect only training costs.The reports show potential usability and harm impacts, so limiting differences to training costs ignores affected-person outcomes.
- The chatbot’s language model must be replaced immediately.Replacement might eventually be considered, but the invalid assumption concerns outcome measurement, not a predetermined technical remedy.
The trap
A successful organizational metric does not automatically demonstrate acceptable human outcomes. How to remember it
A business metric can remain strong while affected employees experience poor usability or harm, so the metric’s sufficiency assumption fails.
How many of these would you get right?
One of 1581 AIGP questions on Certsqill. Take a free five-minute check and see your score per domain — not one number, but which section to open tonight.
Test your AIGP readiness — freeMore Understanding How to Govern AI Development questions
- Risk-acceptance criteria and delegated authority: What should the team establish before deciding whether to →
- Specify an operational human fallback: What should be specified during design? →
- Reassess the affected population and data assumptions: Before relying on the earlier validation, what should →
- All 426 Understanding How to Govern AI Development questions →
Part of the Certsqill AIGP question bank · Understanding How to Govern AI Development ·
Every answer, right and wrong, comes with its own explanation.