WGU D685
Objective Assessment (OA)
Preparation
Practical Applications of Prompt Engineering in AI
50 Original Scenario-Based Questions & Answers with Rationales
2026/2027 Updated | Competency-Weighted | Professionally Formatted
Alignment note
This is an original OA-style preparation set. The competency distribution mirrors the supplied
D685 pre-assessment blueprint: 30% / 24% / 20% / 26%. The questions are newly written
and are not represented as live WGU exam items.
WGU D685 - Practical Applications of Prompt | OA-aligned preparation resource Page 1
, Assessment Blueprint
The set is weighted to the four competency areas shown in the supplied D685 pre-assessment. The
wording is intentionally scenario-focused, with distractors designed to test concept recognition rather than
simple memorization.
Competency Weight Questions
Understands the Importance of Prompt Engineering 30% 1-15 (15 items)
Crafts Effective Prompts 24% 16-27 (12 items)
Evaluates Prompt Mediums 20% 28-37 (10 items)
Performs Prompt Engineering Optimization 26% 38-50 (13 items)
How to use this set
• Attempt all 50 questions before reviewing the answers.
• Pay attention to the exact concept being tested: persona, context, format, bias type, medium, or ethical
principle.
• When two options look plausible, choose the one that most directly addresses the scenario and the
prompt-engineering objective.
• Review every rationale, including items you answered correctly, because the OA can test the same
concept using different wording.
WGU D685 - Practical Applications of Prompt | OA-aligned preparation resource Page 2
, Domain 1 - Understands the Importance of
30%
Prompt Engineering
1. A regional bank wants to review thousands of open-ended customer
comments each week. Which AI capability would provide the greatest
immediate value?
A. Summarizing recurring themes across large volumes of text
B. Replacing all customer-service representatives
C. Guaranteeing that every complaint is factually correct
D. Predicting events that have no relationship to past data
Correct Answer: A. Summarizing recurring themes across large volumes of text
Rationale: Generative AI is well suited to condensing and organizing large amounts of text. It can
surface common themes quickly, while it cannot guarantee correctness or predict unrelated
future events.
2. A manager asks an AI system for a policy recommendation and receives a
confident answer containing a rule that does not exist. What limitation is
illustrated?
A. Faster processing
B. Hallucination
C. Improved personalization
D. Real-time classification
Correct Answer: B. Hallucination
Rationale: A hallucination occurs when a generative model produces plausible-sounding but false
information. Important claims should be verified against authoritative sources.
3. Which feature of a well-designed generative AI interface most directly
helps a user control the response?
A. The ability to remove all uncertainty from the model
B. The ability to replace human judgment
C. The ability to define parameters and constraints
D. The ability to guarantee unbiased output
Correct Answer: C. The ability to define parameters and constraints
Rationale: Controls such as constraints, format requirements, audience, and scope help guide
output. They improve alignment but do not eliminate uncertainty or bias.
WGU D685 - Practical Applications of Prompt | OA-aligned preparation resource Page 3
Objective Assessment (OA)
Preparation
Practical Applications of Prompt Engineering in AI
50 Original Scenario-Based Questions & Answers with Rationales
2026/2027 Updated | Competency-Weighted | Professionally Formatted
Alignment note
This is an original OA-style preparation set. The competency distribution mirrors the supplied
D685 pre-assessment blueprint: 30% / 24% / 20% / 26%. The questions are newly written
and are not represented as live WGU exam items.
WGU D685 - Practical Applications of Prompt | OA-aligned preparation resource Page 1
, Assessment Blueprint
The set is weighted to the four competency areas shown in the supplied D685 pre-assessment. The
wording is intentionally scenario-focused, with distractors designed to test concept recognition rather than
simple memorization.
Competency Weight Questions
Understands the Importance of Prompt Engineering 30% 1-15 (15 items)
Crafts Effective Prompts 24% 16-27 (12 items)
Evaluates Prompt Mediums 20% 28-37 (10 items)
Performs Prompt Engineering Optimization 26% 38-50 (13 items)
How to use this set
• Attempt all 50 questions before reviewing the answers.
• Pay attention to the exact concept being tested: persona, context, format, bias type, medium, or ethical
principle.
• When two options look plausible, choose the one that most directly addresses the scenario and the
prompt-engineering objective.
• Review every rationale, including items you answered correctly, because the OA can test the same
concept using different wording.
WGU D685 - Practical Applications of Prompt | OA-aligned preparation resource Page 2
, Domain 1 - Understands the Importance of
30%
Prompt Engineering
1. A regional bank wants to review thousands of open-ended customer
comments each week. Which AI capability would provide the greatest
immediate value?
A. Summarizing recurring themes across large volumes of text
B. Replacing all customer-service representatives
C. Guaranteeing that every complaint is factually correct
D. Predicting events that have no relationship to past data
Correct Answer: A. Summarizing recurring themes across large volumes of text
Rationale: Generative AI is well suited to condensing and organizing large amounts of text. It can
surface common themes quickly, while it cannot guarantee correctness or predict unrelated
future events.
2. A manager asks an AI system for a policy recommendation and receives a
confident answer containing a rule that does not exist. What limitation is
illustrated?
A. Faster processing
B. Hallucination
C. Improved personalization
D. Real-time classification
Correct Answer: B. Hallucination
Rationale: A hallucination occurs when a generative model produces plausible-sounding but false
information. Important claims should be verified against authoritative sources.
3. Which feature of a well-designed generative AI interface most directly
helps a user control the response?
A. The ability to remove all uncertainty from the model
B. The ability to replace human judgment
C. The ability to define parameters and constraints
D. The ability to guarantee unbiased output
Correct Answer: C. The ability to define parameters and constraints
Rationale: Controls such as constraints, format requirements, audience, and scope help guide
output. They improve alignment but do not eliminate uncertainty or bias.
WGU D685 - Practical Applications of Prompt | OA-aligned preparation resource Page 3
