Loving Change and All things Learning

Module 4: Designing Prompts that Support Authentic Learning

Module 4

Designing Prompts that Support Authentic Learning

Redesigning essay prompts so they draw on course-specific learning, local contexts, and evidence of how students develop their thinking over time.

⚡ Core Principle for Assessment Designers

Generic essay prompts are easier for AI tools to complete because they often rely on widely available information and predictable responses. To restore assessment validity, educators need to create opportunities for students to make decisions, engage with uncertainty, and work through competing perspectives.

1. Why Challenge and Uncertainty Matter in Learning

Most AI tools are designed to simplify tasks and reduce the effort required to produce a response. When applied to essay writing, this “frictionless” automation tempts students to bypass the essential, messy stages of discovery, synthesis, and argument construction.

In contrast, meaningful learning often develops when students encounter complexity, uncertainty, and competing viewpoints. By introducing opportunities for students to draw on personal, local, disciplinary, or course-specific knowledge into our essay prompts, we ensure that outsourcing the task to a chatbot becomes impossible without losing the core substance of the response.

2. Three Features of Effective Assessment Prompts

To design prompts that better reveal student learning and make outsourcing difficult., redesign it using the three dimensions of situatedness:

🏛️ 1. Course-Specific Sources and Experiences

Require students to engage with materials unique to the unit, such as guest lectures, datasets, case studies, seminar discussions, or course resources.. Generic AI tools cannot access unindexed, internal course context.

📍 2. Local & Lived Context

Require students to apply abstract global theories to their local community, workplace environment, or lived experience . This “situated” vantage point cannot be simulated by a database engine.

🔄 3. Reflection on Changing Perspectives

Require students to document an “honest pivot”—explaining what specific piece of evidence changed their initial thesis or forced them to restructure their argument over time.

3. Redesigning Assessment Prompts: Before and After

Examine how generic, AI-vulnerable prompts across different disciplines are more likely to reveal student learning:

Discipline: Political Science & Sociology

❌ BEFORE (Generic & Vulnerable):

“To what extent has social media reshaped how young people participate in political life?”

Flaw: AI generates a balanced, polished 2,000-word essay in 5 seconds using generic metrics (e.g., voting rates, online activism) without any authentic student engagement.

✅ AFTER (Contextualised & Learning-Focused):

“Using the [Local Campus/City Youth Survey Database] analyzed in Week 4, evaluate how local digital campaigns challenge traditional 20th-century metrics of political participation. Contrast these trends with your group’s [‘Cubing’ brainstorm notes] from Week 2 and explain how your initial definition of ‘political engagement’ evolved after encountering conflicting evidence in the [Specific Seminar Reading].”

Why it works: Requires unindexed course data, personal brainstorming notes, and explicit documentation of a change in thinking.

Discipline: History & Urban Studies

❌ BEFORE (Generic & Vulnerable):

“Analyze the primary factors contributing to social exclusion in modern urban public spaces.”

✅ AFTER (Contextualised & Learning-Focused):

“Using the 1984 [City Council] Parks Committee Minutes from our private course archive, evaluate how physical park bench design choices conflict with David Harvey’s global theory of ‘The Right to the City’. Your response must cite specific local architectural features and explain how your argument shifted when you encountered challenges raised by the primary source.”
Discipline: Business, Management & Healthcare

❌ BEFORE (Generic & Vulnerable):

“Discuss the ethical challenges of implementing artificial intelligence in organizational management.”

✅ AFTER (Contextualised & Learning-Focused):

“Draw on your own workplace experience or a local community organization to evaluate an AI adoption policy [1, 13]. Contrast official management goals with the insights shared by our [Week 6 Guest Lecturer], and include a 300-word reflection detailing how you resolved conflicting evidence in your final recommendations .”

4. Academic Prompt Redesign Checklist

Use this quick audit checklist when drafting or revising unit assignment briefs:

📋 PROMPT REDESIGN AUDIT CHECKLIST

[ ] 1. Internal Course Anchor: Does the prompt require direct engagement with an unindexed course resource (e.g., private archive, specific lecture transcript, seminar dataset)?

[ ] 2. Local / Lived Application: Does the prompt require students to ground abstract theories in a specific local community, workplace, or personal context?

[ ] 3. Multiple Perspectives: Does the prompt require students to compare different viewpoints, interpretations, data sources, or experiences?

[ ] 4. Development of Thinking: Does the task require students to show how their ideas changed, developed, or became more refined over time?

[ ] 5. AI Input Testing: Have you run the prompt through ChatGPT or Claude yourself to verify that a standard, generic response fails the specific disciplinary criteria?