- Unit
- Unit 4 of 20
- Recommended period
- Week 3
- Primary reading
- Textbook Chapter 4
- Estimated study
- 7–9 hours
- Prerequisites
- Units 1–3
- Required platform
- None
- Primary framework
- PCAFE-V
- Assessment connection
- Assignment 2 starts
Recommended first action: Select one weak prompt from a supplied example or your own low-risk
practice. Save the original before revising it so Assignment 2 can show an authentic development process.
Begin the prompt diagnostic
Unit overview
Unit 4 introduces purposeful prompt design as a structured communication and workflow task. You will learn to
define the human purpose before composing the prompt, separate task instructions from context and sources,
specify the intended audience and output format, establish observable evaluation criteria, and identify what
must be verified independently. The unit introduces the author-approved PCAFE-V framework and emphasizes that a
detailed prompt may improve relevance without guaranteeing accuracy, fairness, safety, authorization, or
suitability.
Purpose and task
Define the problem, intended outcome, AI role, scope, and constraints before requesting an output.
Context and audience
Provide only necessary, authorized information and adapt the requested output to the intended reader or
user.
Evaluation and verification
Specify observable quality criteria and identify consequential claims requiring independent evidence.
Central idea: An effective prompt improves the conditions for a
useful provisional output. Human users remain responsible for context, sources, evaluation, verification,
decisions, and consequences.
Unit learning outcomes
By the end of Unit 4, you should be able to:
- Distinguish human purpose, task, prompt, context, source, audience, format, criterion, and verification
requirement.
- Diagnose a weak prompt by identifying missing purpose, audience, constraints, evidence, and review
expectations.
- Write a bounded task statement that defines what the AI may and may not do.
- Select only the context necessary and authorized for the legitimate purpose.
- Specify an audience and output format appropriate to the intended use.
- Develop observable evaluation criteria for relevance, completeness, accuracy, structure, accessibility, and
usability.
- Distinguish output evaluation from independent verification.
- Identify ambiguity, assumptions, missing information, conflicts, and conditions requiring clarification.
- Apply the author-approved PCAFE-V framework to design and revise a prompt.
- Compare initial and revised prompts and outputs using evidence rather than preference alone.
- Document the prompt-development process for Assignment 2.
- Identify when prompting is unsuitable and a human-led or non-AI approach should be used.
Unit 4 lessons
Complete the five lessons in order. These filenames are local to the Unit 4 folder.
Lesson 1 of 5 · Suggested time: 70–85
minutes
4.1 Purpose, Task, and Prompt Boundaries
Define the human purpose, intended outcome, scope, constraints, AI role, human-only activities, and
conditions requiring clarification or stopping.
Lesson 2 of 5 · Suggested time: 70–85
minutes
4.2 Context, Sources, Privacy, and Assumptions
Separate necessary from unnecessary context, establish source boundaries, minimize sensitive information,
and make assumptions and missing information explicit.
Lesson 3 of 5 · Suggested time: 70–85
minutes
4.3 Audience, Format, and Accessible Output
Specify intended users, language, structure, length, tone, accessibility, and output constraints without
treating style as proof of quality.
Lesson 4 of 5 · Suggested time: 75–90
minutes
4.4 Evaluation Criteria and Verification Requirements
Develop observable criteria, identify claims and calculations requiring evidence, and define reviewer,
escalation, and stop conditions.
Lesson 5 of 5 · Suggested time: 90–120
minutes
4.5 PCAFE-V Prompt Redesign and Comparison
Apply the author-approved PCAFE-V framework, compare initial and revised results, document corrections, and
preserve the evidence required for Assignment 2.
Folder convention: Keep unit-04.html and t-01.html
through t-05.html together in a dedicated Unit 4 folder or Moodle content package.
Embedded Unit 4 study guide
This sequence supports one week of asynchronous study and begins the Prompt Engineering Portfolio.
- Diagnose
Identify what a weak prompt fails to define.
- Design
Add purpose, context, audience, format, and criteria.
- Test
Generate or inspect a provisional output and verify important content.
- Compare
Explain which revisions helped, which did not, and why.
- Session 1: Retrieval and weak-prompt diagnosis, 60–75
minutes.
Retrieve Unit 3 suitability principles, read Chapter 4, and annotate a weak prompt.
- Session 2: Lesson 1, 70–85 minutes.
Define purpose, task, scope, constraints, AI
role, human-only activities, and stopping.
- Session 3: Lesson 2, 70–85 minutes.
Minimize context, define sources, remove
identifying details, and surface assumptions.
- Session 4: Lesson 3, 70–85 minutes.
Specify audience, format, language,
accessibility, length, and structure.
- Session 5: Lesson 4, 75–90 minutes.
Build evaluation and verification requirements
and identify qualified reviewers.
- Session 6: Lesson 5 and portfolio preparation, 90–120 minutes.
Apply PCAFE-V,
compare versions, document the revision rationale, and update Assignments 1 and 2.
Evidence to retain
- Original weak prompt and diagnosis
- Purpose and task boundary
- Context-minimization record
- Audience and format specification
- Evaluation and verification checklist
- PCAFE-V prompt
- Initial and revised outputs or supplied alternatives
- Comparison and correction record
- Assignment 1 reflection notes
Show the prompt-development sequence
- Confirm that AI is suitable for the human purpose.
- Define the task, AI role, scope, and constraints.
- Provide only necessary and authorized context.
- Identify permissible sources and excluded sources.
- Specify the audience and output format.
- Define observable evaluation criteria.
- State verification requirements and responsible reviewers.
- Ask for clarification when essential information is missing.
- Evaluate, verify, correct, and compare.
- Document the process, human decisions, and remaining uncertainty.
Assessment milestones in Unit 4
Assignment 2: Prompt Engineering Portfolio
Weight: 15%
Start: Unit 4, Week 3
Develop: Across Units 4–6
Recommended submission: End of Module 2, Week 4
Official Moodle due date and time: ____________________
Assignment 1: Journal continues
Record how defining purpose, minimizing context, specifying criteria, and requiring verification changed
the output and your responsibilities.
Recommended final submission: End of Module 6, Week 11
Official Moodle due date and time: ____________________
Preserve the process: Assignment 2 should show prompt development.
Save the initial prompt, output, diagnosis, revised prompt, revised output, evaluation, verification, and
explanation of changes.
Show the Unit 4 portfolio-record checklist
- Task and legitimate purpose
- Original prompt and output
- Weakness diagnosis
- PCAFE-V redesign
- Context and source decisions
- Audience and format
- Evaluation criteria
- Verification performed
- Corrections and comparison
- Final human decision and required disclosure
Required learning activities
Unit 4 activities and evidence
| Activity |
Purpose |
Expected product |
Estimated time |
| Weak-prompt diagnosis |
Identify missing purpose, context, audience, format, criteria, and verification. |
Annotated prompt |
30–45 minutes |
| Purpose and task statement |
Define legitimate purpose, scope, AI role, constraints, and human-only activities. |
Task brief |
25–35 minutes |
| Context-minimization exercise |
Separate necessary, optional, sensitive, and prohibited information. |
Context inventory |
30–45 minutes |
| Audience and format design |
Specify intended reader, language, structure, accessibility, and output limits. |
Output specification |
25–35 minutes |
| Evaluation and verification plan |
Define quality criteria, evidence requirements, reviewers, and escalation. |
Review checklist |
35–50 minutes |
| PCAFE-V redesign |
Transform the weak prompt into a complete purpose-driven prompt. |
Revised prompt |
40–55 minutes |
| Prompt comparison |
Compare initial and revised outputs using observable evidence. |
Comparison record |
45–60 minutes |
Participation options
Option A: Approved AI system
Test prompts in an approved system using synthetic, public, supplied, or authorized information.
Option B: Supplied records
Use supplied prompts, outputs, context inventories, and revision records.
Option C: Non-AI analysis
Design, critique, and compare task instructions without operating an AI platform.
All options assess the same prompt-design, evaluation, verification, and responsibility outcomes.
Responsible-use boundaries and support
Boundaries for Unit 4
- Do not enter confidential, personal, proprietary, privileged, security-sensitive, or culturally governed
information.
- Do not assume that more context is always better.
- Do not request fabricated citations, concealed AI use, impersonation, or unauthorized perspectives.
- Do not treat a detailed prompt or polished output as verified.
- Do not delegate approval, consequential decisions, or accountability to AI.
Support and accessibility
- Use supplied prompts and outputs if live AI is inaccessible, unavailable, or unwanted.
- Specify accessible output requirements where relevant.
- Contact the course coordinator for assignment or permitted-use questions.
- Use library, accessibility, privacy, or technical support as needed.