(PBL-SGT)-Fusion
(PBL-SGT)-Fusion is an active-learning method that adapts each session to the learners actually in the room: problem-based learning fused with small-group teaching, with a pre-test diagnostic before every session so the difficulty matches the group rather than an imagined average student. The name and the method are from my MMedSci thesis. The result that mattered: when teaching adapts, learners outperform their stage.
Where it came from
From 2013 to 2018 I was a Clinical Teaching Fellow, teaching University of Nottingham medical students at Lincoln County Hospital, and that classroom became my laboratory. The method took shape there and became my MMedSci thesis, ‘A study of combining active learning techniques within problem based learning when teaching medical students in a small group environment’ (Medical Education, Distinction, University of Nottingham, 2015). Here is the naming as the thesis states it:
The term I have applied to this process is (PBL-SGT)-Fusion, as I have taken elements from traditional Problem Based Learning (PBL) and Small Group Teaching (SGT) when developing this style.
I evaluated it with psychometrics, item difficulty and score reliability, and with Kirkpatrick’s framework, across 35 students (15 junior-phase, 20 senior-phase). The thesis closed by proposing a follow-on model, ‘Theory, Simulation, Practise’ (TSP). The full 146-page thesis is self-hosted, with its abstract, at the research page.
The argument
The ward had stopped being a reliable classroom. As the thesis frames it, the European Working Time Directive replaced the old firm structure with shift work and transient teams, so the environment students were being prepared for was also the environment degrading their teaching. My answer was to make classroom time work harder: a short pre-test at the start of each session to activate prior knowledge and show the tutor the group’s real baseline, the session pitched to that baseline rather than to a syllabus average, a group case discussion mapped as it happened, and a post-test under examination conditions to consolidate and catch gaps.
It worked better than I dared hope. Junior-phase (CP1) students mastered material written for senior-phase (CP3) students, and the junior cohort’s median post-test mark was higher than the senior cohort’s pre-test mark. When teaching adapts, learners outperform their stage.
This was adaptive learning built by hand, a decade before AI made personalisation cheap. The principle transfers directly: the pre-test diagnostic is what a good adaptive system does continuously, and what a good teacher does instinctively. Two years after the thesis I was diagnosed with dyslexia. The diagnosis did not create the method; it explained it. I had spent years building around a way of learning I had not yet named in myself.
Who it is for
Teachers first, in any discipline: the method asks for nothing exotic, a pre-test, a tutor who reads it, and the willingness to change the plan for the room. But it is also for anyone building adaptive or AI-personalised learning today. The by-hand version is evidence that the principle, adapt to the learner in front of you, predates the tooling; the tools have finally caught up with the pedagogy. That transfer is my argument, not the thesis’s finding; the ledger below says so.
What (PBL-SGT)-Fusion does not claim
A promising result is not a licence to overclaim it, so let me be precise about the edges.
- It does not claim external validation. As the thesis itself says, the technique had been implemented by one teacher, at one site, with 35 students; it has not been externally replicated.
- It does not claim problem-based learning or small-group teaching as mine. The fusion of the two, and the pre-test diagnostic that drives the adaptation, are what the name covers.
- It does not claim long-term outcomes. The evidence is written-exam performance and reported enjoyment within one placement, from a convenience sample.
- It does not claim AI validation. The thesis predates the AI era; the transfer of its principle to AI-personalised learning is my hypothesis.
The claims, classed
I sort what I publish by how strong the claim is, so here is this page graded by my own rubric. The classes are Observation, Hypothesis, Coined term, Internal result and Externally validated finding; each claim names its evidence and where that evidence comes from.
The ledger
The term itself, (PBL-SGT)-Fusion. Coined term. Evidence: the thesis, formal title above, examined and awarded with Distinction in 2015. Source: owner-supplied document, self-hosted in full at the research page; the degree award is the institutional check.
The method can be delivered, and it improves written-exam performance. Internal result, institutionally examined. Evidence: the thesis, 35 students, evaluated with psychometrics (item difficulty, score reliability) and Kirkpatrick’s framework. Source: owner-supplied document; the degree award is the institutional check, not external replication.
Junior-phase students mastered senior-phase material, with a higher median post-test mark than the senior cohort’s pre-test. Internal result, from the same study. Source: owner-supplied document; the same institutional check.
Students of all stages received it well and asked for more. Internal result, self-reported by the students within the same study. Source: owner-supplied document.
The adaptive principle transfers to AI-era personalised learning. Hypothesis. Evidence: the argument on this page. Source: my own proposed transfer, not yet validated; this page will say so until that changes.
Watch and read
- The thesis itself: (PBL-SGT)-Fusion, MMedSci, University of Nottingham, 2015, with its abstract and the full 146-page PDF
- The teaching practice around it: Education, five years as a Clinical Teaching Fellow and WatMed Educational
- The reflection where the method met its explanation: Education fails when it cannot adapt to the learner
- The human arc: My story, the classroom that became a laboratory