AML Communication & Stakeholder Relations 2 — Questions and Answers
Question 1: A data scientist presents model performance metrics to a marketing team unfamiliar with ML. Which visualization approach is most effective?
- Show ROC curves and AUC scores directly
- Translate metrics into business outcomes like revenue impact or customer retention rates (Correct answer)
- Present confusion matrices with all technical details
- Use mathematical formulas to explain precision and recall
Correct answer: Translate metrics into business outcomes like revenue impact or customer retention rates
Translating technical metrics into business outcomes makes model performance meaningful and actionable for non-technical stakeholders.
Question 2: When a key stakeholder requests a feature that would compromise model fairness, what is the best initial response?
- Implement the feature immediately to maintain the relationship
- Refuse the request without explanation
- Acknowledge the request, explain fairness implications with data, and propose compliant alternatives (Correct answer)
- Escalate to legal without informing the stakeholder
Correct answer: Acknowledge the request, explain fairness implications with data, and propose compliant alternatives
Acknowledging the request while presenting evidence of fairness risks and offering alternatives preserves relationships while upholding ethical standards.
Question 3: Which technique is most appropriate for communicating model uncertainty to executive stakeholders?
- Report only point estimates to avoid confusion
- Present confidence intervals alongside predictions with plain-language explanations (Correct answer)
- Provide full probability distributions with statistical notation
- Omit uncertainty metrics to appear more confident
Correct answer: Present confidence intervals alongside predictions with plain-language explanations
Confidence intervals paired with plain-language explanations help executives understand prediction reliability without overwhelming technical detail.
Question 4: A stakeholder insists that a model's accuracy is 'not good enough' without defining a threshold. What should the ML professional do?
- Keep improving the model indefinitely
- Facilitate a requirements workshop to define acceptable performance metrics and business thresholds (Correct answer)
- Switch to a completely different algorithm
- Present the model as satisfactory without further discussion
Correct answer: Facilitate a requirements workshop to define acceptable performance metrics and business thresholds
Facilitating a structured discussion to define measurable acceptance criteria aligns stakeholder expectations with technical realities.
Question 5: When communicating the limitations of a deployed ML model to end users, which approach best balances transparency and usability?
- Hide all limitations to maximize adoption
- Provide plain-language disclosures about known failure modes and appropriate use cases (Correct answer)
- Share full technical documentation with every user
- Only disclose limitations if legally required
Correct answer: Provide plain-language disclosures about known failure modes and appropriate use cases
Plain-language disclosures about known failure modes empower users to apply models appropriately and build justified trust.
Question 6: During a sprint review, a business stakeholder asks why the model improved on validation data but not in production. What is the most effective response?
- Promise the issue will be fixed next sprint without explanation
- Explain the concept of distribution shift using a concrete analogy relevant to the business domain (Correct answer)
- Defer the question to the next technical meeting
- Increase model complexity to close the gap
Correct answer: Explain the concept of distribution shift using a concrete analogy relevant to the business domain
Using domain-relevant analogies to explain distribution shift bridges the gap between technical causes and business understanding.
Question 7: A cross-functional team disagrees on model deployment timelines due to differing risk tolerances. What communication strategy should the ML lead use?
- Force a technical decision without input from other teams
- Facilitate a structured risk assessment discussion using a shared framework to align all stakeholders (Correct answer)
- Delay deployment indefinitely to avoid conflict
- Let each team implement their preferred timeline independently
Correct answer: Facilitate a structured risk assessment discussion using a shared framework to align all stakeholders
A shared risk assessment framework creates common ground and objective criteria that help cross-functional teams reach consensus on deployment timing.
AML Communication & Stakeholder Relations 2
A data scientist presents model performance metrics to a marketing team unfamiliar with ML.
Which visualization approach is most effective?