Frame the driving question and inquiry arc

1

Research

Why this matters, and what good looks like

A driving question frames the problem students investigate. PBLWorks expects a meaningful question at the right level of challenge. HQPBL adds sustained work on a complex issue with more than one possible answer. AI can generate contrast and flag wording against criteria you supply. Only you can judge authenticity, cultural fit and whether students can investigate the question in your setting.

Keep the learning intent and authentic context fixed before generating wording.
Require more than one defensible response supported by evidence.
Reject questions a quick search can settle.
Make the required subject knowledge necessary to a strong answer.
Test the question against real time, resources and students.

These sources support PBL question design. They do not establish that AI improves question quality or saves time.

Where the evidence comes from

Strong for PBL question design; AI benefit unvalidated.

2

Workflow

Brief it, steer it, check it

Ask for real contrast

Fill the context with your own project. Remove names, survey responses and personal data before copying anything. Several angles give you something to compare and reject.

Prompt 1 · Generate

CONTEXT
Subject and year group: [add them].
Fixed learning intent and priority standards: [paste them].
Authentic context, audience or problem: [describe it].
Class-level starting knowledge, without names or personal data: [add it].
Time, resources and other constraints: [add them].

REQUEST
Write five contrasting driving questions for this project.
Make the angles genuinely different.

QUALITY BAR
Each question must allow more than one defensible response.
A quick search must be insufficient.
A strong answer must require the learning intent and standards above.
Do not invent an audience, partner or local fact.

FORMAT
For each option, give the question, its angle and what students would need to investigate.

Steer the nearest option

Choose the nearest miss and name the exact weakness. Keep the useful part while changing the part that fails.

Prompt 2 · Steer

CONTEXT
Chosen candidate question: [paste it].
What must stay: [name it].
Fixed constraints that must appear verbatim in every revision: [paste them].
Exact weakness: [name it].
Learning intent and authentic context: [paste them].

REQUEST
Write three revisions that fix the named weakness in different ways.

QUALITY BAR
Keep the stated strength and repeat every fixed constraint verbatim.
Do not lower the challenge, change the scope of a number or add an unconfirmed claim.
Each revision must still need the required subject learning.

FORMAT
For each revision, show the question, the change made and the remaining risk.

Critique before choosing

Use your criteria to find weaknesses. Read the output as a prompt to look harder, not as a verdict.

Prompt 3 · Critique

CONTEXT
Candidate driving question: [paste it].
Priority standards and learning intent: [paste them].
Authentic context: [paste it].
Project length and constraints: [add them].
Teacher quality criteria: [paste them].

REQUEST
Critique the question against the teacher quality criteria supplied above.

QUALITY BAR
Use evidence from the supplied context for each judgement.
Do not rewrite the question.
Reject the candidate before building an inquiry arc if a quick search can answer it, it is detached from the standards or its authenticity is unconfirmed.
Mark anything the context cannot prove as a teacher check.

FORMAT
Use a table with criterion, pass or concern, evidence and teacher decision needed. Include one explicit line for each stop condition: answerable by a quick search, detached from the standards, and authenticity unconfirmed. For each line, record the evidence, reject or teacher-check decision, and any unresolved check. Do not claim a search was performed unless search evidence is supplied; otherwise mark the quick-search test as a teacher check. Then draft a three-sentence rationale using only passed evidence and naming unresolved checks.

Build the inquiry arc

Once you approve the question, work backwards from what students need to know, decide and revise.

Prompt 4 · Inquiry arc

CONTEXT
Teacher-approved driving question: [paste it].
Teacher decisions on stop conditions: [record evidence and decisions for quick-search sufficiency, standards connection and confirmed authenticity; leave unresolved checks on hold].
Priority standards and learning intent: [paste them].
Project length, resources and fixed milestones: [add them].
Known starting points: [add them].

REQUEST
Draft the inquiry arc that could support students in answering the question.

QUALITY BAR
Include sub-questions, knowledge-building needs, decision points and revision triggers.
Do not assume resources, partners or prior knowledge that are absent from the context.
Mark every assumption for teacher review. If no assumption is found, write none identified and name the supplied source for the entry.

FORMAT
Use a table with sequence, sub-question, knowledge or skill, decision point, revision trigger and assumption or source.

Check before you use it

Could a capable student answer the question in ten minutes online?
Can two well-supported answers genuinely disagree?
Does a strong answer require the priority subject learning?
Is the context authentic and confirmed?
Can your students investigate it with the time and resources available?
Does the inquiry arc build knowledge before major decisions?

Stop rule

Reject questions answerable by a quick search, detached from standards or falsely presented as authentic.

3

Reuse it

So the next one takes minutes

Suitable

Build a skill

Save this as a reusable instruction you can run on any project.

The generation, steering and critique sequence stays stable across projects.

Not suitable

Build an agent

Chain the steps, with a checkpoint where you approve.

Question quality depends on a bounded teacher judgement, so autonomous follow-up adds little value.