Design discussion questions and follow-up prompts

Author: AILesson8 min setupTested with:ChatGPTReviewed: 2026-08-28

Quick answer

Build an evidence-centered question sequence that opens participation and deepens reasoning. Provide: Discussion material, Learning goals and boundaries, Group and facilitation context. Expected result: A timed discussion guide with question purposes, follow-ups, participation moves, and evidence checks.

Add your context

Your text stays in this browser. AILesson Prompts does not send it to a model or server.

Discussion materialLearning goals and boundariesGroup and facilitation context

Your prompt

Unfilled fields remain visible as placeholders, so you can still copy and edit the prompt

Design a discussion-question sequence grounded in the supplied material.

Source material, positions, evidence, and locators:
[material]

Interpretation, comparison, evaluation, decision goals, and scope boundaries:
[goals]

Learner level, group size, time, prior knowledge, participation, access, and sensitivity:
[setting]

Create fewer, deeper questions in a purposeful arc: entry/observation, interpretation, evidence, comparison, assumption, counterexample, implication, and synthesis or decision only where aligned. Each main question must be open enough for more than recall yet specific enough to answer from the material. Avoid leading premises, false balance, double-barreled questions, forced personal disclosure, speculative claims about real people, and questions with a single hidden instructor answer.

For each question provide purpose, source anchor, expected reasoning—not a required opinion—wait time, likely productive paths, misconception or facilitation risk, and follow-ups for evidence, clarification, challenge, connection, and extension. Add participation moves for silence, dominant voices, disagreement, and remote/access needs without assigning identities. Include what the facilitator should not resolve, a timed run-of-show with removable questions, exit evidence, and an alignment/bias audit.

Private by defaultPrompt assembly happens locally in your browser. Avoid placing confidential information into any AI service unless your organization allows it.

From input to outcome

A worked example

See how concrete context turns this recipe into a usable result

Actual input

Discussion material: An anonymized library case: prior six weeks 107/198 reservations collected before expiry (54%); reminder-pilot six weeks 131/214 (61%). Periods were not randomized and seasonal title mix differed. Eleven patrons opted out, three reported late messages, staff handled replies about 25 minutes/day. A proposed randomized test has no approved sample rationale, owner, budget, accessibility/privacy review, or stop thresholds.
Learning goals and boundaries: Graduate public-service students should distinguish description from causation, weigh service benefit against operational/equity guardrails, compare design/refine/decline options, and state conditions for a proportionate decision. Permanent rollout and guessing individual patron motives are out of scope.
Group and facilitation context: Twenty-four students, 35 minutes, groups of four then whole class. Students know percentages but not trial design. A few voices often dominate. One remote participant and bilingual English-Chinese key terms. Topic includes privacy but no personal disclosure should be requested.

Example output

0-4 min silent entry, then round-robin: Q1 “List two statements the numbers directly support and one they do not.” Purpose: claim boundary; anchors 107/198, 131/214, non-randomization. Wait 45 seconds. Evidence follow-up: “Which exact feature prevents the causal statement?” Risk: treating seven points as caused effect. Do not lecture before all groups state a boundary.

4-11 min Q2: “What plausible differences between the periods could change how we interpret 54% and 61%, and which are actually supplied?” Purpose: distinguish evidence from hypotheses. Follow-ups: “Is that in the case or a verification need?” “What observation would separate those explanations?” Do not ask students to guess patron motives.

11-19 min Q3: “How should opt-outs, late messages, and staff minutes change the design decision?” Paths may frame measures, thresholds, access, or capacity. Challenge: “Could the same fact support different choices under different risk tolerances?” Avoid assuming any threshold. Q4 if time: “Whose interests are affected, and what evidence—not presumed views—would be needed?” No personal privacy stories required.

19-29 min groups choose Design, Refine, or Decline. Q5: “State the strongest evidence for your option, its strongest weakness, and two conditions that would change it.” Assign rotating roles: evidence locator, alternative challenger, recorder, reporter; remote participant can take any role via shared text. Dominant voice move: each person writes before open discussion; disagreement move: paraphrase the competing reasoning before rebuttal.

29-34 whole-class comparison; Q6: “Which conditions are necessary before design work, and which only before test launch?” Facilitator should not declare one option correct where conditions differ. Remove Q4 if behind, not Q5. 34-35 exit: one bounded conclusion plus one missing fact. Audit: all questions serve causal reasoning, guardrails, alternatives, or proportionality; no forced disclosure, false balance, or invented actor belief; bilingual terms and spoken/text participation support access.

Why this works

1
A question arc moves discussion from noticing evidence to evaluating implications instead of jumping to opinions
2
Planned follow-ups help the facilitator deepen reasoning without steering learners toward a preferred answer

Check the result

Can each main question be answered from the material through multiple defensible reasoning paths
Do follow-ups request evidence or reasoning without embedding the desired conclusion
Does the facilitation plan protect participation, disagreement, time, and access

Use it with confidence

Frequently asked questions

Practical answers about when to use this recipe, what to provide, and where human review still matters

What should I prepare before using “Design discussion questions and follow-up prompts”?

For “Design discussion questions and follow-up prompts,” prepare Discussion material, Learning goals and boundaries, and Group and facilitation context. Replace placeholders only with information you can verify. If a detail is unknown, preserve that uncertainty explicitly instead of asking the model to infer it.

When is the “Design discussion questions and follow-up prompts” result not ready to use?

The result is not ready if it does not yet deliver the stated outcome—A timed discussion guide with question purposes, follow-ups, participation moves, and evidence checks—from the supplied evidence, or if it relies on unresolved assumptions, missing approvals, or invented details. Use the checks as release gates: revise the source inputs or assign a named, authorized reviewer instead of polishing an unsupported output.

Which AI tools have recorded tests for “Design discussion questions and follow-up prompts”?

The published test record for “Design discussion questions and follow-up prompts” lists ChatGPT as of 2026-08-28. This confirms recorded runs, not guaranteed compatibility or identical results in later product versions. For another tool or version, keep every constraint visible and repeat the result checks before use.

More ways to explore