WGU D685: Practical Applications of Prompt
Complete Practice Package • 122 Original OA-Style Questions • Answers & Rationales Immediately After Each Question
Important: These are original practice questions aligned to WGU's published D685 course description and prompt-engineering
concepts. They are not reproduced or represented as confidential/leaked WGU assessment questions.
1. A marketing analyst asks an AI system to summarize a 30-page report in five bullet points for an
executive audience. Which prompt element most directly specifies the desired presentation?
A. Persona
B. Output format
C. Context
D. Model architecture
Correct Answer: B. Output format
Rationale: Output format tells the model how the response should be structured, such as bullet points, a table, or a
specified length.
2. A user writes, “Explain photosynthesis.” The response is too broad. Which change most directly
improves specificity?
A. Ask for a definition in one sentence for a ninth-grade audience
B. Remove the subject from the prompt
C. Ask the model to use no context
D. Increase the number of unrelated examples
Correct Answer: A. Ask for a definition in one sentence for a ninth-grade audience
Rationale: Adding audience and scope constraints makes the task more specific and reduces ambiguity.
3. A manager asks an AI tool to “write an email.” Which addition best provides context?
A. Use exactly 120 words
B. The recipient is a customer whose order was delayed two days
C. Use a friendly tone
D. Return the answer as plain text
Correct Answer: B. The recipient is a customer whose order was delayed two days
Rationale: Context supplies relevant background facts that help the model tailor its response.
4. Which prompt is most likely to produce a consistent structured response?
A. Tell me about project risks.
B. Discuss project risks however you want.
C. List five project risks in a table with columns for risk, likelihood, impact, and mitigation.
D. Project risks?
Correct Answer: C. List five project risks in a table with columns for risk, likelihood, impact, and mitigation.
Rationale: The third prompt specifies the task, quantity, and output structure, making the requested result more
constrained and repeatable.
WGU D685 — Original Practice Questions Page 1
,5. A user asks an AI system to critique a résumé and says, “Act as a career coach.” What prompt
component is being supplied?
A. Persona
B. Output format
C. Training data
D. Tokenization
Correct Answer: A. Persona
Rationale: A persona establishes the perspective or role the model should adopt while completing the task.
6. A prompt says, “Use the attached policy document and identify three requirements that apply to
remote employees.” Which part supplies the source material?
A. Context
B. Randomness
C. Model weights
D. Temperature
Correct Answer: A. Context
Rationale: The referenced policy document provides contextual information the model should use for the task.
7. A user gives an AI system one example of an input paired with the desired output before asking it to
solve a second similar problem. What technique is this?
A. Zero-shot prompting
B. One-shot prompting
C. Fine-tuning
D. Reinforcement learning
Correct Answer: B. One-shot prompting
Rationale: One-shot prompting provides one example to demonstrate the desired pattern.
8. A user provides several examples of how customer complaints should be categorized before asking
the model to classify a new complaint. What technique is this?
A. Few-shot prompting
B. Zero-shot prompting
C. Unsupervised learning
D. Retrieval-only prompting
Correct Answer: A. Few-shot prompting
Rationale: Few-shot prompting uses multiple examples to guide the model toward the desired pattern.
WGU D685 — Original Practice Questions Page 2
,9. A user asks an AI model to perform a task without providing examples. What is this called?
A. Few-shot prompting
B. One-shot prompting
C. Zero-shot prompting
D. Transfer learning
Correct Answer: C. Zero-shot prompting
Rationale: Zero-shot prompting asks the model to perform the task without demonstrations.
10. A generated answer contains a plausible citation that does not exist. What AI behavior does this
illustrate?
A. Hallucination
B. Compression
C. Tokenization
D. Overfitting
Correct Answer: A. Hallucination
Rationale: A hallucination is generated content that is plausible but unsupported, false, or fabricated.
11. A student asks an AI model to answer a factual question and then independently checks the result
against a reliable source. What practice is being demonstrated?
A. Output validation
B. Prompt deletion
C. Model pretraining
D. Data poisoning
Correct Answer: A. Output validation
Rationale: Validation means checking generated output against authoritative or otherwise appropriate evidence.
12. A prompt asks, “Give me three causes of the problem and explain the evidence for each.” Which
instruction primarily encourages analytical depth?
A. The word 'three'
B. The request to explain evidence
C. The word 'problem'
D. The use of quotation marks
Correct Answer: B. The request to explain evidence
Rationale: Requiring evidence encourages the model to support each cause rather than simply list possibilities.
WGU D685 — Original Practice Questions Page 3
, 13. A user repeatedly changes a prompt after reviewing the model’s output until the answer meets the
task requirements. What process is this?
A. Prompt iteration
B. Model pretraining
C. Data labeling
D. Token pruning
Correct Answer: A. Prompt iteration
Rationale: Prompt iteration is the process of refining prompts based on observed output quality.
14. An AI response is accurate but much too detailed for a quick executive briefing. Which prompt
revision is most direct?
A. Request a 75-word executive summary with five bullets
B. Ask for more technical jargon
C. Remove the audience
D. Add unrelated background information
Correct Answer: A. Request a 75-word executive summary with five bullets
Rationale: A clear length and format constraint directly targets the excessive detail.
15. A user wants an AI system to generate a comparison of two products. Which instruction best
reduces ambiguity?
A. Compare them.
B. Tell me everything about them.
C. Compare battery life, warranty, price, and repairability in a four-column table.
D. Which one is better?
Correct Answer: C. Compare battery life, warranty, price, and repairability in a four-column table.
Rationale: Specifying comparison criteria and a table format defines the task and output structure.
16. Which statement best describes prompt engineering?
A. Changing the model’s underlying weights
B. Designing and refining inputs to guide a model toward useful outputs
C. Replacing a database with a chatbot
D. Eliminating the need for human review
Correct Answer: B. Designing and refining inputs to guide a model toward useful outputs
Rationale: Prompt engineering focuses on constructing and refining inputs that guide generative AI behavior.
WGU D685 — Original Practice Questions Page 4
Complete Practice Package • 122 Original OA-Style Questions • Answers & Rationales Immediately After Each Question
Important: These are original practice questions aligned to WGU's published D685 course description and prompt-engineering
concepts. They are not reproduced or represented as confidential/leaked WGU assessment questions.
1. A marketing analyst asks an AI system to summarize a 30-page report in five bullet points for an
executive audience. Which prompt element most directly specifies the desired presentation?
A. Persona
B. Output format
C. Context
D. Model architecture
Correct Answer: B. Output format
Rationale: Output format tells the model how the response should be structured, such as bullet points, a table, or a
specified length.
2. A user writes, “Explain photosynthesis.” The response is too broad. Which change most directly
improves specificity?
A. Ask for a definition in one sentence for a ninth-grade audience
B. Remove the subject from the prompt
C. Ask the model to use no context
D. Increase the number of unrelated examples
Correct Answer: A. Ask for a definition in one sentence for a ninth-grade audience
Rationale: Adding audience and scope constraints makes the task more specific and reduces ambiguity.
3. A manager asks an AI tool to “write an email.” Which addition best provides context?
A. Use exactly 120 words
B. The recipient is a customer whose order was delayed two days
C. Use a friendly tone
D. Return the answer as plain text
Correct Answer: B. The recipient is a customer whose order was delayed two days
Rationale: Context supplies relevant background facts that help the model tailor its response.
4. Which prompt is most likely to produce a consistent structured response?
A. Tell me about project risks.
B. Discuss project risks however you want.
C. List five project risks in a table with columns for risk, likelihood, impact, and mitigation.
D. Project risks?
Correct Answer: C. List five project risks in a table with columns for risk, likelihood, impact, and mitigation.
Rationale: The third prompt specifies the task, quantity, and output structure, making the requested result more
constrained and repeatable.
WGU D685 — Original Practice Questions Page 1
,5. A user asks an AI system to critique a résumé and says, “Act as a career coach.” What prompt
component is being supplied?
A. Persona
B. Output format
C. Training data
D. Tokenization
Correct Answer: A. Persona
Rationale: A persona establishes the perspective or role the model should adopt while completing the task.
6. A prompt says, “Use the attached policy document and identify three requirements that apply to
remote employees.” Which part supplies the source material?
A. Context
B. Randomness
C. Model weights
D. Temperature
Correct Answer: A. Context
Rationale: The referenced policy document provides contextual information the model should use for the task.
7. A user gives an AI system one example of an input paired with the desired output before asking it to
solve a second similar problem. What technique is this?
A. Zero-shot prompting
B. One-shot prompting
C. Fine-tuning
D. Reinforcement learning
Correct Answer: B. One-shot prompting
Rationale: One-shot prompting provides one example to demonstrate the desired pattern.
8. A user provides several examples of how customer complaints should be categorized before asking
the model to classify a new complaint. What technique is this?
A. Few-shot prompting
B. Zero-shot prompting
C. Unsupervised learning
D. Retrieval-only prompting
Correct Answer: A. Few-shot prompting
Rationale: Few-shot prompting uses multiple examples to guide the model toward the desired pattern.
WGU D685 — Original Practice Questions Page 2
,9. A user asks an AI model to perform a task without providing examples. What is this called?
A. Few-shot prompting
B. One-shot prompting
C. Zero-shot prompting
D. Transfer learning
Correct Answer: C. Zero-shot prompting
Rationale: Zero-shot prompting asks the model to perform the task without demonstrations.
10. A generated answer contains a plausible citation that does not exist. What AI behavior does this
illustrate?
A. Hallucination
B. Compression
C. Tokenization
D. Overfitting
Correct Answer: A. Hallucination
Rationale: A hallucination is generated content that is plausible but unsupported, false, or fabricated.
11. A student asks an AI model to answer a factual question and then independently checks the result
against a reliable source. What practice is being demonstrated?
A. Output validation
B. Prompt deletion
C. Model pretraining
D. Data poisoning
Correct Answer: A. Output validation
Rationale: Validation means checking generated output against authoritative or otherwise appropriate evidence.
12. A prompt asks, “Give me three causes of the problem and explain the evidence for each.” Which
instruction primarily encourages analytical depth?
A. The word 'three'
B. The request to explain evidence
C. The word 'problem'
D. The use of quotation marks
Correct Answer: B. The request to explain evidence
Rationale: Requiring evidence encourages the model to support each cause rather than simply list possibilities.
WGU D685 — Original Practice Questions Page 3
, 13. A user repeatedly changes a prompt after reviewing the model’s output until the answer meets the
task requirements. What process is this?
A. Prompt iteration
B. Model pretraining
C. Data labeling
D. Token pruning
Correct Answer: A. Prompt iteration
Rationale: Prompt iteration is the process of refining prompts based on observed output quality.
14. An AI response is accurate but much too detailed for a quick executive briefing. Which prompt
revision is most direct?
A. Request a 75-word executive summary with five bullets
B. Ask for more technical jargon
C. Remove the audience
D. Add unrelated background information
Correct Answer: A. Request a 75-word executive summary with five bullets
Rationale: A clear length and format constraint directly targets the excessive detail.
15. A user wants an AI system to generate a comparison of two products. Which instruction best
reduces ambiguity?
A. Compare them.
B. Tell me everything about them.
C. Compare battery life, warranty, price, and repairability in a four-column table.
D. Which one is better?
Correct Answer: C. Compare battery life, warranty, price, and repairability in a four-column table.
Rationale: Specifying comparison criteria and a table format defines the task and output structure.
16. Which statement best describes prompt engineering?
A. Changing the model’s underlying weights
B. Designing and refining inputs to guide a model toward useful outputs
C. Replacing a database with a chatbot
D. Eliminating the need for human review
Correct Answer: B. Designing and refining inputs to guide a model toward useful outputs
Rationale: Prompt engineering focuses on constructing and refining inputs that guide generative AI behavior.
WGU D685 — Original Practice Questions Page 4
