Module 2 • Prompting

Module 2: AI Prompting Lab

Iterate prompts (v1–v3), compare outputs side-by-side, apply a critique checklist, and document a fact-check source.

Page image
Prompt template

Use this structure

Role: You are a critical-thinking partner.

Context:
- Situation: [paste your TEL / Design Thinking problem]
- Constraints: [time, tools, stakeholders, ethics]

Task:
1) Reframe the problem in 3 perspectives (learner, instructor, system/risk).
2) Generate 3 alternative strategies (Plan A/B/C) with trade-offs.
3) Provide an evidence checklist: what must be verified and how.
4) State uncertainty + assumptions explicitly.

Then critique my approach and propose a Plan B under a new hypothesis.
Critique checklist
Tick items after reviewing the AI output.
Evidence-first rule
Treat AI output as a hypothesis. Verify claims using a credible source.
Prompting guidance aligns with higher-ed recommendations emphasizing clarity, specificity, iterative refinement, and verification (see Faculty Focus).
Side-by-side compare

Prompt versions (v1 → v3)

Write prompts and paste outputs. Compare improvements across versions.

Exports
v1 — baseline
v1
v2 — constrained
v2
v3 — verified
v3
Critique checklist
Select what your prompts/outputs achieved.
Fact-check panel
Reflection
Reflection quality rubric
Claim → Evidence → Alternative → Limitation
exemplars
For this module, refer to how you reframed the problem, what constraints you added, and what evidence you used to verify or reject the AI output.
What strong reflections include
  • Claim: your current hypothesis or decision.
  • Evidence: cite the log/source/observation that supports it.
  • Alternative: at least one competing explanation.
  • Limitation: what you still don’t know + next test.
Common weak patterns
  • Vague actions (“I tried something”).
  • No evidence cited.
  • No Plan B or alternative frame.
  • Overconfidence; no uncertainty.

I made the prompt better. The answer looked good so I used it.

Claim: v3 is more reliable because it forces evidence checks and uncertainty.
Evidence: v1 lacked constraints; v2 added counterarguments; v3 required citations, failure modes, and a Plan B.
Alternative: AI could still hallucinate even with citations, so I verified the key claim against an external source and noted remaining uncertainty.
Limitation + next test: I would cross-check with a second source and test the recommendation in a low-risk environment.
Inline rubric scoring

Module 2: AI Prompting Lab — FT rubric entry

1–4

Score how well your prompting process supported reframing, alternatives, evidence checks, and reflective monitoring (1–4).

Reframing & Perspective Shift
Did you shift frames (technical / learning-design / risk) when needed?
Alternative Generation & Strategy Switching
Did you generate and test alternatives (Plan A/B/C) instead of sticking to one path?
Evidence-Based Justification
Did you cite concrete evidence (logs/sources/observations) to justify decisions?
Reflection & Transfer
Did you articulate what transfers and propose a Plan B for new contexts?
—
Preview

Rubric radar

This radar is generated from your rubric entry in this browser.
Tips
  • Anchor scores to concrete evidence (log line, source link, or observation).
  • Include at least one alternative frame or Plan B.
  • Use the exports in Tools & Evaluation for your thesis snapshots.