Training · Cross-cuttingMoveprovenverified

Problem-Based Learning

also known as PBL, Barrows PBL, case-based PBL, inquiry-based problem learning

Tags: probleminquirysmall-groupill-structuredreasoning

Learners encounter a complex, ill-structured real-world problem before they have been taught the content needed to solve it. The problem — not the instructor's lecture — drives what gets learned. Working in small facilitated groups, learners identify what they know, what they do not know, and what they need to find out, then pursue that knowledge and return to apply it. The method was pioneered by Howard S. Barrows and Robyn M. Tamblyn at McMaster University's medical school in the 1960s. Their 1980 book formally defined it as 'the learning that results from the process of working toward an understanding or resolution of the problem.' In AI and tech training, problem-based learning appears as scenario-driven cohort sessions. A broken system, a failing model, or a contested business decision is the opening trigger rather than a slide deck.

How the learner advances

Intent. Force learners to build the knowledge they need by confronting an ill-structured real problem before they have the answers — making the acquisition of content purposeful rather than preparatory.

When to apply. Apply when the target competency is reasoning under uncertainty, not recall of procedures. Use problem-based learning when the domain involves complex, ambiguous situations with no single correct answer. The learner must need to develop diagnostic or clinical reasoning — in the broad sense, not just medicine. The cohort must be able to sustain small-group work with a facilitator who withholds answers. The timeline must allow multiple inquiry-and-return cycles. Avoid when learners lack the baseline vocabulary to engage with the problem at all — a very brief direct-instruction primer may be needed first.

Threshold — earns the next step. The learner can work through a novel problem in the same domain independently, generating hypotheses, identifying their own knowledge gaps, pursuing self-directed research, and applying new knowledge back to the problem — without waiting for an instructor to structure the inquiry.

Masterpiece — the artifact that proves it. The group's learning synthesis document: a written or presented account of the hypotheses they generated, the knowledge gaps they identified, the resources they used, the conclusions they reached, and the reasoning moves they now have available for future problems.

Facets

  • Containerworkshop
  • Modeappliedcollaborativeself-directed
  • Reachcohort
  • Personadeveloperanalyst-opsnon-technical
  • Craft (AI Fluency)discernmentsynthesiscollaboration
  • Learnerhuman
  • Trainerhuman

Inputs

  • Trigger problem or caseA realistic, ill-structured scenario that is complex enough to reveal genuine knowledge gaps but scoped enough to be tractable in the available time. The problem should not have a single correct answer retrievable by lookup.
  • Small learning group (5–9 learners)A group small enough for every member to be heard and accountable, large enough for productive diversity. Barrows' original tutorials at McMaster used groups of five to nine students.
  • Facilitator who does not answerA tutor or coach who guides group process, surfaces hidden assumptions, and asks deepening questions — but who does not supply the content answers the group is seeking. The facilitator's restraint is what makes the problem-driven inquiry work.

Outputs

  • A more capable learnerA learner who has practised clinical or diagnostic reasoning — the ability to move from incomplete information to a working hypothesis, test it, and revise. This reasoning skill transfers across problems in the domain.
  • Problem resolution and learning synthesisThe group's documented working-through of the problem: what they hypothesised, what they sought, what they found, and how their understanding shifted. This synthesis — the masterpiece — is the evidence of learning, not the final answer.

Steps (5)

  1. Encounter the problem

    The group receives the trigger material — a patient case, a failing system log, a contested business decision — before any instruction. They read or observe it together. The facilitator does not explain what the problem is asking.

  2. Generate hypotheses and identify learning issues

    The group brainstorms possible explanations for the problem, identifies what they know that is relevant, and explicitly lists what they do not know — the learning issues. These learning issues become the self-study agenda.

  3. Self-study and resource acquisition

    Each learner independently researches their assigned or chosen learning issues, using whatever sources are available. This phase is genuinely self-directed — the facilitator does not assign readings.

  4. Return, share, and apply

    The group reconvenes, shares what each member found, integrates the new knowledge, and applies it back to the original problem. Hypotheses are revised, rejected, or confirmed. A new trigger or added information may deepen the case.

  5. Reflect and abstract

    The group reflects on both the problem resolution and the learning process: what reasoning moves worked, what gaps remain, what they would do differently next time. The facilitator draws out transferable principles.

Principles

  • The problem precedes the instruction — knowledge acquired in response to a felt need sticks; knowledge acquired in preparation for a hypothetical future problem does not.
  • Ill-structure is a feature, not a bug — a problem with a single correct retrievable answer trains lookup, not reasoning. The ambiguity is what activates clinical thinking.

Known uses (2)

Known failure modes (3)

Related trainings (3)

Sources (2)

Provenance

  • Ecosystem: education
  • Added to catalog:
  • Last updated:
  • Verification status: verified