Quiz: Module 3 - Crafting Effective Prompts
Test your understanding with 30 multiple-choice questions.
- 27-30 correct: Expert level!
- 21-26 correct: Proficient
- 15-20 correct: Review recommended
- Below 15: Please review the module
Questions 1-10: Clarity and Specificity
Question 1
What is the main problem with vague prompts?
- A) They cost more tokens
- B) They lead to vague, unpredictable responses
- C) They are faster to process
- D) They work better with GPT-4
View Answer
Correct Answer: B
Vague prompts lead to vague responses because the AI has to guess what you want.
Question 2
What does the "5 W's + H" framework help with?
- A) Counting words
- B) Ensuring prompt completeness (What, Who, Where, When, Why, How)
- C) Measuring response quality
- D) Calculating token costs
View Answer
Correct Answer: B
The 5 W's + H (What, Who, Where, When, Why, How) ensures your prompt addresses all necessary dimensions.
Question 3
Which is a more specific alternative to "Write something short"?
- A) "Write something brief"
- B) "Write a short piece"
- C) "Write 100-150 words"
- D) "Write less"
View Answer
Correct Answer: C
Quantifying with exact numbers ("100-150 words") removes ambiguity.
Question 4
What should you do with vague words like "good" or "detailed"?
- A) Use them frequently
- B) Replace them with concrete descriptions
- C) Capitalize them
- D) Put them in quotes
View Answer
Correct Answer: B
Replace vague words with concrete descriptions. Instead of "detailed," specify "including statistics, quotes, and case studies."
Question 5
What is the "abstraction ladder" concept?
- A) A coding framework
- B) Moving from vague to concrete language
- C) A type of prompt structure
- D) A way to count tokens
View Answer
Correct Answer: B
The abstraction ladder helps you move from abstract (vague) to concrete (specific) language.
Question 6
Which technique helps ensure completeness in prompts?
- A) Using shorter prompts
- B) Specifying what to include AND exclude
- C) Avoiding all constraints
- D) Using only questions
View Answer
Correct Answer: B
Specifying both inclusions and exclusions ensures the AI covers what you want and avoids what you don't.
Question 7
What is a "success criteria" in prompt engineering?
- A) How many tokens were used
- B) Measurable standards to evaluate if output meets requirements
- C) The model's accuracy score
- D) The speed of response
View Answer
Correct Answer: B
Success criteria are specific, measurable standards that define what a good output looks like.
Question 8
How should you handle numbers in prompts?
- A) Avoid them entirely
- B) Use words instead of digits
- C) Be precise and quantify when possible
- D) Round to the nearest hundred
View Answer
Correct Answer: C
Be precise with numbers: "exactly 5 examples" instead of "some examples."
Question 9
What does the clarity checklist help with?
- A) Counting words
- B) Verifying prompt clarity before sending
- C) Measuring AI performance
- D) Calculating costs
View Answer
Correct Answer: B
The clarity checklist helps verify that your prompt is unambiguous and complete.
Question 10
"Give me a few examples" should be replaced with:
- A) "Give me examples"
- B) "Give me exactly 3 examples"
- C) "Give me some examples"
- D) "Give me multiple examples"
View Answer
Correct Answer: B
Specific numbers ("exactly 3 examples") eliminate ambiguity.
Questions 11-20: Context and Constraints
Question 11
What does context do for AI responses?
- A) Makes them shorter
- B) Transforms generic responses into relevant ones
- C) Increases token usage only
- D) Slows down processing
View Answer
Correct Answer: B
Context helps the AI understand the situation and produce more relevant, tailored responses.
Question 12
What are the four types of context?
- A) Short, Medium, Long, Extended
- B) Role, Audience, Purpose, Situational
- C) Text, Code, Data, Image
- D) Input, Output, Process, Memory
View Answer
Correct Answer: B
The four types are Role (who), Audience (for whom), Purpose (why), and Situational (background).
Question 13
What is the context window?
- A) A visual interface
- B) Maximum text the AI can process at once
- C) A type of prompt format
- D) The editing window
View Answer
Correct Answer: B
The context window is the maximum amount of text (prompt + response) an AI model can process at once.
Question 14
What happens when you provide too many constraints?
- A) Better responses
- B) Faster processing
- C) Over-restricted, potentially awkward outputs
- D) Lower costs
View Answer
Correct Answer: C
Too many constraints can over-restrict the AI, leading to awkward or impossible outputs.
Question 15
Which is a good length constraint?
- A) "Make it short"
- B) "Between 150-200 words"
- C) "Not too long"
- D) "Average length"
View Answer
Correct Answer: B
Specific ranges ("150-200 words") are clear and actionable.
Question 16
What is Role Context?
- A) The AI's job title
- B) Defining who the AI should be when responding
- C) Your role at work
- D) The format of output
View Answer
Correct Answer: B
Role context defines the persona the AI should adopt (e.g., "You are a senior financial analyst...").
Question 17
How can you save tokens while providing context?
- A) Avoid all context
- B) Use summaries instead of full documents
- C) Only use context for long prompts
- D) Repeat information multiple times
View Answer
Correct Answer: B
Using summaries of key points instead of full documents preserves information while reducing tokens.
Question 18
What should LOW CONTEXT + LOW CONSTRAINTS prompts be avoided?
- A) They're too expensive
- B) They lead to unpredictable results
- C) They're too slow
- D) They only work with Claude
View Answer
Correct Answer: B
Low context + low constraints = unpredictable results because the AI has little guidance.
Question 19
What is Audience Context?
- A) How many people will read it
- B) Defining who will receive/read the output
- C) The social media platform
- D) The file format
View Answer
Correct Answer: B
Audience context defines who the output is for (demographics, knowledge level).
Question 20
Which model has the largest context window (approximately)?
- A) GPT-3.5 (4K)
- B) GPT-4 (128K)
- C) Claude 3 (200K)
- D) All are equal
View Answer
Correct Answer: C
Claude 3 has the largest context window at approximately 200K tokens.
Questions 21-30: Formatting and Structure
Question 21
Why is markdown useful in prompts?
- A) It's required by all AI models
- B) It helps organize prompts and specify output formatting
- C) It reduces token count
- D) It only works with code
View Answer
Correct Answer: B
Markdown helps organize complex prompts and request specific formatting in outputs.
Question 22
What is an output template?
- A) A predefined structure for the AI to fill in
- B) A type of AI model
- C) A programming language
- D) A measurement of quality
View Answer
Correct Answer: A
An output template is a predefined structure that the AI fills in, ensuring consistent outputs.
Question 23
What is a benefit of using templates?
- A) Slower responses
- B) Consistency and completeness
- C) Higher costs
- D) Less control
View Answer
Correct Answer: B
Templates ensure consistency, completeness, parseability, and professionalism.
Question 24
When should you request JSON output?
- A) Never
- B) When you need to programmatically process the data
- C) Only for creative writing
- D) To save tokens
View Answer
Correct Answer: B
JSON is ideal when you need structured data that can be programmatically processed.
Question 25
What is layered detail in formatting?
- A) Adding more words
- B) Providing explanations at multiple complexity levels
- C) Using multiple fonts
- D) Adding colors
View Answer
Correct Answer: B
Layered detail provides explanations at different complexity levels (child, student, expert).
Question 26
What should a code output specification include?
- A) Just the language name
- B) Function name, parameters, return type, requirements
- C) Only comments
- D) No specifications needed
View Answer
Correct Answer: B
Good code specifications include function name, parameters, return type, requirements, and style expectations.
Question 27
What is a reusable prompt framework?
- A) A one-time prompt
- B) A template structure that can be adapted for similar tasks
- C) A type of AI model
- D) A measurement tool
View Answer
Correct Answer: B
A reusable framework is a template structure with placeholders that can be adapted for similar tasks.
Question 28
Which markdown element is used for emphasis?
- A) # Headers
- B) Bold and italic
- C) - Bullet points
- D) | Tables |
View Answer
Correct Answer: B
Bold (double asterisks) and italic (single asterisks) are used for emphasis.
Question 29
How do you request a table in output?
- A) Just say "make a table"
- B) Specify rows, columns, and what each should contain
- C) Tables can't be generated
- D) Use JSON instead
View Answer
Correct Answer: B
Specify the structure: rows (what to compare), columns (dimensions), and any additional notes.
Question 30
What is multi-part response structuring?
- A) Using multiple AI models
- B) Dividing output into labeled sequential sections
- C) Writing prompts in parts
- D) Sending multiple requests
View Answer
Correct Answer: B
Multi-part structuring divides the output into clearly labeled sections (Part 1, Part 2, etc.).
Score Summary
| Score | Level | Recommendation |
|---|---|---|
| 27-30 | Expert | Proceed to Module 4 |
| 21-26 | Proficient | Proceed, review weak areas |
| 15-20 | Developing | Review before proceeding |
| 0-14 | Beginning | Please review Module 3 |