AI PBL Bootcamp · Free · Three Mondays from 26 October
Improve

Sequence projects across years and subjects

Projects repeat content, overload certain periods or fail to build prerequisite knowledge, independence and complexity over time.

AI does

Cross-document comparison and gap visualization.

You do

Judge progression, developmental appropriateness and what should be removed or deferred.

1

Research

Why this matters, and what good looks like

A progression map shows how later projects build on earlier knowledge and practices. Here you review one link across two years and two subjects, not a whole curriculum. PBLWorks offers professional teaching guidance; the EEF handbook offers evaluation guidance. Their support for this sequencing task is indirect. Neither validates cross-year prerequisites, developmental progression or timetable optimisation. AI benefit remains unvalidated.

  • PBLWorks asks teachers to plan for their context and students.
  • PBLWorks connects support with project goals and assessment of knowledge and understanding. Exposure alone cannot establish mastery.
  • EEF asks evaluators to examine implementation context and reasons for adaptations.

The bounded comparison and decision records below are local process choices. Teachers must establish subject-specific validity themselves.

Where the evidence comes from

Moderate for curriculum coherence; AI benefit unvalidated.

2

Workflow

Brief it, steer it, check it
  1. Bound the progression link

    Choose two projects across two years and subjects. Bring standards, project plans and timetable extracts. Optional curriculum maps and reusable project packs can supply these; equivalent teacher documents work. Remove learner data before using AI.

    Prompt 1 · Inventory
    CONTEXT
    Scope and teacher-checked source extracts with IDs, versions, exact standards, concepts, practices, assessment requirements and timetable constraints: [sources].
    Private human checks summarised without learner data, including readiness evidence limits: [checks].
    REQUEST
    Inventory this single progression link and flag repetition, gaps and conflicting requirements for teacher review.
    QUALITY BAR
    Use only supplied text. Never infer mastery from exposure or equate standards or subjects. Keep readiness judgements attributed to teachers. Missing or stale standards, readiness checks or timetable constraints mean HOLD; do not draft a sequence. Do not grade or request individual evidence.
    FORMAT
    Table: Item ID | Year and subject | Source ID and version | Exact evidence | Gap or repetition | Status.
    End with Human checks needed. Use HOLD or Supplied only for Status.
  2. Compare bounded options

    Open the originals and validate the inventory with both subject teachers. Record corrections and prerequisite judgements. A prerequisite is knowledge needed before the next task. If checks remain unresolved, stop this link rather than guessing.

    Prompt 2 · Options
    CONTEXT
    Exact approved inventory: [inventory].
    Versioned teacher corrections and checks against original standards, subject concepts, readiness and timetable: [decisions].
    REQUEST
    Propose two options for this link, preserving the distinct requirements of each subject. Identify what repeats intentionally and what changes in independence or complexity.
    QUALITY BAR
    Apply only explicit teacher edits. Preserve source IDs and evidence limits. Transfer means applying learning in a different context, not automatic competence. Do not select an option, remove required content or invent capacity. Missing human validation means HOLD and no options. No institutional commitment.
    FORMAT
    Table: Option ID | Sequence | Prerequisites and readiness | Subject-specific transfer | Planned variation | Timetable tradeoff | Evidence IDs | Status.
    Use O1 and O2 when options are possible; Status is Proposal or HOLD. End with Teacher choice needed.
  3. Record the selected map

    Teachers choose, remove or defer the link after checking subject accuracy and feasibility. Record their decision separately. Keep assessment evidence private; give AI only non-identifying requirements and the limits of the readiness judgement.

    Prompt 3 · Map
    CONTEXT
    Exact approved options: [options].
    Versioned teacher selection, edits and confirmed requirements, including evidence limits: [selection].
    REQUEST
    Build the selected two-project progression map. Include standards, concepts, practices, prerequisites, transfer, assessment evidence requirements and planned variation.
    QUALITY BAR
    A proposal is never approval. Use only the explicit selection and edits. Attribute verification to the supplied human record, not your own review. Keep unknowns HOLD. Do not infer mastery, interchange subjects, grade, or turn a proposed timetable into an institutional commitment. Without selection return HOLD only.
    FORMAT
    Table: Project ID | Year and subject | Standards and concepts | Practices | Prerequisites | Transfer | Assessment evidence required | Planned variation | Timetable | Evidence IDs | Status.
    End with Deferred or unresolved items. Status is Teacher-selected proposal or HOLD.
  4. Audit the teacher-selected result

    Teachers check the map against originals and record edits before this audit. They decide actual use afterwards. Stop and re-scope if active work exceeds 60 minutes; retain unresolved links instead of presenting a full curriculum.

    Prompt 4 · Audit
    CONTEXT
    Exact map and versioned teacher edits: [map].
    Source extracts and versioned human verification record, including the selected option: [verification].
    REQUEST
    Audit the teacher-selected map for unsupported standards, sequence, readiness, transfer, assessment and timetable claims. Do not rewrite it.
    QUALITY BAR
    Quote exact map text and compare only with supplied evidence. Label supplied human decisions separately from verified facts. Missing, stale or conflicting checks mean HOLD. Stop if exposure becomes mastery or subjects are interchangeable. Never approve release or claim educational impact.
    FORMAT
    Table: Check | Exact map quote | Evidence ID | Finding | Human action.
    Cover Scope, Standards, Readiness, Transfer, Assessment, Timetable. End with Unresolved holds and Human release remains required.

Check before you use it

  • Have both subject teachers checked the exact standards and subject-specific concepts against originals?
  • Is readiness a bounded human judgement supported by assessment, rather than inferred from exposure?
  • Does the map retain each subject's requirements and test transfer separately?
  • Have teachers checked timetable feasibility and kept unconfirmed commitments on HOLD?
  • Are the chosen variation, evidence requirements and deferred links traceable to explicit teacher decisions?
  • Have you removed learner data and kept release with the responsible teachers?

Stop rule

Stop if AI infers mastery from exposure or treats standards/subjects as interchangeable.

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.

A bounded comparison can reuse its structure with fresh teacher checks.

Not suitable

Build an agent

Chain the steps, with a checkpoint where you approve.

Consequential decisions and private verification remain with humans; autonomous monitoring is outside scope.

Help us improve this method

Anonymous, and it takes under a minute.

Did you get something usable?
How useful was this method?
1 = wasted my time · 5 = I'll use it again
Where did it break down?

Please don't include names, school names or anything about individual students.

No account, no email, nothing stored about you.

Use cases are starting points with one perspective. Disagree with the process, that's expected. Make it yours.

Co-funded by the European Union under Erasmus+ KA210-SCH. No student data. No teacher material stored.