Chapter 35 Quiz: Change Management for AI

Chapter 35 Quiz: Change Management for AI

Multiple Choice

Question 1. According to the chapter, what percentage of Athena's regional managers were overriding the demand forecasting model's recommendations before the change management program was implemented?

(a) 22%
(b) 48%
(c) 68%
(d) 82%

Question 2. In the ADKAR model, which element was identified as the most critical gap in Athena's initial demand forecasting rollout?

(a) Awareness — managers did not know the model existed
(b) Desire — incentive structures discouraged adoption, and Ability — poor workflow integration
(c) Knowledge — managers did not know what AI is
(d) Reinforcement — managers were punished for using the model

Question 3. Which of the following best describes the McKinsey ratio for AI program investment?

(a) For every dollar spent on AI, organizations should spend one dollar on data
(b) For every dollar spent on AI technology, three to five dollars should be spent on change management, training, and process redesign
(c) AI technology accounts for 60-70% of total program costs
(d) Change management should consume no more than 10% of the AI budget

Question 4. What distinguishes AI-driven organizational change from previous technology waves, according to the chapter?

(a) AI is more expensive than previous technologies
(b) AI replaces judgment rather than just tasks, is probabilistic rather than deterministic, is more opaque, and triggers identity-level concerns
(c) AI requires less training than previous technologies because it is more intuitive
(d) AI only affects technical employees, unlike previous technologies which affected everyone

Question 5. In Kotter's 8-step model, which step does the chapter identify as most often skipped in AI transformations?

(a) Step 1: Create a Sense of Urgency
(b) Step 3: Develop a Vision and Strategy
(c) Step 6: Generate Short-Term Wins
(d) Step 8: Anchor New Approaches in the Culture

Question 6. A research study cited in the chapter found that the strongest predictor of AI resistance was:

(a) Fear of job loss (economic anxiety)
(b) Identity threat — the perception that AI diminishes the value of one's professional expertise
(c) Lack of technical literacy
(d) Organizational distrust of management

Question 7. The "last mile" problem in AI refers to:

(a) The difficulty of training the final layer of a deep neural network
(b) The gap between a technically deployed model and a model that is actually used by its intended users
(c) The final stage of data pipeline integration
(d) The last phase of model evaluation before production deployment

Question 8. A 2024 Gartner study found that approximately what percentage of AI projects that reach production fail to deliver their expected business value?

(a) 25%
(b) 50%
(c) 65%
(d) 85%

Question 9. Which of the following was NOT identified as a common "last mile" failure mode?

(a) The Dashboard Nobody Opens
(b) The Recommendation Ignored
(c) The Algorithm Too Accurate
(d) The Expert Override

Question 10. The chapter argues that giving users the ability to override the model has what paradoxical effect?

(a) It decreases model accuracy but increases user satisfaction
(b) It increases adoption because employees who feel forced to follow recommendations resent the loss of autonomy
(c) It eliminates the need for change management
(d) It makes the model unnecessary over time

Question 11. In the centaur model of human-AI collaboration, at which level does Athena's post-bias-crisis HR department operate?

(a) Level 1: AI Decides, Human Monitors
(b) Level 2: AI Recommends, Human Decides
(c) Level 3: AI Assists, Human Leads
(d) Level 4: Human Decides, AI Learns

Question 12. What was the completion rate difference between Athena's in-person AI literacy training and the e-learning pilot?

(a) In-person: 72%, E-learning: 45%
(b) In-person: 85%, E-learning: 60%
(c) In-person: 94%, E-learning: 31%
(d) In-person: 99%, E-learning: 50%

Question 13. In Athena's workforce impact assessment of 12,000 employees, approximately what percentage of roles were classified as Zone 3 (Transitional — genuinely at risk of elimination)?

(a) 1%
(b) 8%
(c) 15%
(d) 34%

Question 14. Which of the following is a leading indicator (rather than a lagging indicator) of AI adoption success?

(a) Sustained usage over six months
(b) Business outcome improvement
(c) Manager communication quality about AI in team meetings
(d) Override rate trends

Question 15. What was Ravi's view on the optimal override rate for the demand forecasting model?

(a) 0% — managers should always follow the model
(b) 5% — overrides should be rare exceptions
(c) 15-25% — low enough to capture model value, high enough to incorporate human judgment
(d) 50% — an even split between human and AI decision-making

Short Answer

Question 16. Explain why Ravi did not mandate compliance with the demand forecasting model's recommendations, despite the model being more accurate than the managers. Reference at least two concepts from the chapter in your answer.

Question 17. Describe how Athena repositioned the AI tools in its HR department following the bias crisis of Chapter 25. What change in the human-AI collaboration level did this represent, and why was it necessary for rebuilding trust?

Question 18. The chapter states that "resistance is information, not obstruction." Explain this statement using two specific examples from the Athena case — one where resistance revealed a legitimate data problem and one where it revealed a change management gap.

Question 19. Compare the communication needs of executives, middle managers, and frontline employees when an AI initiative is announced. Identify one message element that should be emphasized for each group and one mistake to avoid for each group.

Question 20. Explain the concept of "just-in-time learning" as applied to AI adoption. Why is it more effective than front-loaded training for helping employees develop new AI-related skills? Reference the forgetting curve in your answer.

Scenario-Based Questions

Question 21. A logistics company deploys an AI-powered route optimization system. Delivery drivers report that the system sometimes recommends routes through neighborhoods they know to be difficult (narrow streets, construction zones, unreliable parking). The system is 15% more efficient overall, but drivers are overriding it on 40% of routes.

(a) Which ADKAR element is the primary bottleneck?
(b) Name two structural changes the company should implement.
(c) What is the risk of mandating compliance in this situation?

Question 22. A pharmaceutical company's AI drug discovery team has developed a model that predicts which molecular compounds are most likely to succeed in clinical trials. The model has been validated and shows promising results. However, senior research scientists with decades of experience are resistant, arguing that "chemistry is not a pattern-matching exercise."

Using the resistance patterns from Section 35.4, identify which pattern(s) are most likely at work and propose a change management approach that addresses the scientists' concerns without diminishing the AI system's potential contribution.

Question 23. A retail bank has deployed an AI chatbot for customer service. In the first month, customer satisfaction scores drop from 4.2/5 to 3.6/5, and call center volume increases by 12% as customers who interact with the chatbot call the help line for follow-up.

(a) Is this a technical failure, a change management failure, or both? Explain.
(b) Using the last mile framework, identify at least two failure modes that might explain these results.
(c) Propose three corrective actions, prioritized by impact.

Answer key for multiple-choice questions and guidance for short-answer questions are available in Appendix B.