It surveys how this term is defined in the published literature — what the sources agree on, where they diverge and what remains contested — and carries its own numbered bibliography. It does not describe what we run. Programme policy is on the core pages.
Our own use of the word is often narrower than general usage; the short definition and the boundary are in the glossary, under Problem-based learning. The review date below matters, because the literature moves.
A definitional review. Citations follow IEEE style; see References.
Abstract
Problem-Based Learning is an instructional method in which small groups learn by working through an ill-structured problem under a facilitator, identifying what they do not know and researching it themselves. It originated in medical education and remains concentrated there. It shares the acronym PBL with Project-Based Learning, an unrelated pedagogy from progressive school reform; this article covers the problem sense only, and the companion review covers the other. Set out here are the definition, the lineage, the features that separate it from its two neighbours, what four decades of research report, and the claims that research does not support.
I. Definition
The most cited working definition in the educational-psychology literature is Hmelo-Silver’s:
Problem-based learning (PBL) is an instructional method in which students learn through facilitated problem solving. In PBL, student learning centers on a complex problem that does not have a single correct answer. Students work in collaborative groups to identify what they need to learn in order to solve a problem. They engage in self-directed learning (SDL) and then apply their new knowledge to the problem and reflect on what they learned and the effectiveness of the strategies employed. The teacher acts to facilitate the learning process rather than to provide knowledge [1, p. 235].
Savery’s overview, the standard reference in the method’s own journal, defines it as “an instructional (and curricular) learner-centered approach that empowers learners to conduct research, integrate theory and practice, and apply knowledge and skills to develop a viable solution to a defined problem” [2].
Five goals are conventionally attributed to the method: constructing an extensive and flexible knowledge base; developing effective problem-solving skills; developing self-directed, lifelong learning skills; becoming an effective collaborator; and becoming intrinsically motivated to learn [1, p. 240].
The mechanism is a cycle rather than a syllabus. A group is presented with minimal information about a complex problem; it identifies the relevant facts, generates hypotheses, and — critically — identifies its own knowledge deficiencies, which become the learning issues researched during self-directed study. The group then reconvenes, applies what it found, revises its hypotheses, and reflects on both the problem and its own process [1, pp. 236–242]. A four-column whiteboard — Facts, Ideas, Learning Issues, Action Plan — externalises the cycle and is the method’s characteristic tool [1, p. 242].
II. Origins
Problem-Based Learning was developed in the late 1960s to reform medical education at McMaster University in Canada, and further refined at Maastricht University in the Netherlands from 1974 [5]. The canonical statement of the approach is Barrows and Tamblyn’s 1980 book [3]; the central tenet usually quoted from it is that “the problem comes first” — the problem precedes rather than applies the teaching [4]. That book was not obtained for this review and is cited here for attribution only; the quotation is reproduced as it appears in [4].
A bibliometric review of 14,130 Scopus-indexed documents dates the first conceptualisation and implementation efforts to medical schools in the Netherlands, Canada and the USA “during the 1970s and 1980s”, and attributes the design of PBL-oriented curricula to Maastricht in the 1970s [4]. The same review finds the literature’s growth “low and flat” through the 1970s and 1980s, with 58% of all PBL publications appearing between 2010 and 2019 [4].
The intellectual context is broader than medicine. Problem-based approaches sit within the experiential tradition of Kilpatrick and Dewey and alongside case-based instruction and project-based learning [1, p. 236] — which is exactly why the boundaries need stating.
III. What distinguishes it
A. From Project-Based Learning — the acronym collision
The two pedagogies have different origins, different units of work and different assessment logic. Project-Based Learning descends from Kilpatrick’s project method of 1918 and was developed for science education by Blumenfeld and colleagues in the 1990s; Problem-Based Learning descends from McMaster in 1969 [5]. The operational contrast, as coded in a meta-analysis that treated them as separate methods [5]:
| Problem-Based | Project-Based | |
|---|---|---|
| Unit of work | one ill-structured problem, usually two sessions days or a week apart | a project running weeks or months |
| Role of the problem | starts the learning; no preparation beforehand | provides the context for a driving question |
| Product | optional, in the reporting phase | always culminates in a tangible product or presentation |
| Student control | medium–high; the tutor checks that student-set learning goals match the intended objectives | high; the process is student-driven |
The assessment consequence is the sharpest of these. Problem-based work is assessed on the reasoning a group performs and on what its members learn; project-based work is assessed on a sustained artefact. It has been argued that the distinction is worth defending “since some approaches to project-based learning, unlike problem-based learning, require that teachers specify the final product and provide guidance on the students’ approach to creating the product” [7, citing 2].
The boundary is not universally accepted. Thomas took the view that problem-based learning studies “have all of the defining features” of project-based learning — centrality, driving question, constructive investigation, autonomy and realism — differing only in a tutorial ingredient, and included them in his project-based review [6]. A later review took the opposite course, excluding problem-based learning as a “close cousin” whose definitional overlap made pooling unsafe [7]. Both positions are defensible; what is not defensible is using “PBL” without saying which is meant.
B. From case-based learning — a second collision
“CBL” is ambiguous too. In the motivation meta-analysis it abbreviates case-based learning, not challenge-based learning [5]. Case-based learning differs from problem-based learning on timing and autonomy: cases are prepared in advance and presented at the end of the learning process to apply knowledge already taught, the questions are usually predetermined, and post-session self-study is uncommon [5]. Where problem-based learning puts the problem first, case teaching applies theory learned earlier through lecture and discussion [4].
C. What is actually load-bearing
Hmelo-Silver compares problem-based learning against anchored instruction and project-based science on the type and role of the problem, the process, the teacher’s role, collaboration and tools — and concludes that “the SDL emphasis is a distinguishing feature of PBL” [1, pp. 238–239]. In the neighbouring approaches the teacher does some direct instruction when the problem-solving demands it; in problem-based learning, students become responsible for identifying and closing their own knowledge gaps [1, p. 239]. That, rather than realism or group work, is the discriminating feature.
On the external-stakeholder axis, problem-based learning is internally set and internally judged: the problem is authored for its pedagogical yield, and the tutor checks the group’s self-set learning goals against the curriculum’s intended objectives [5]. It is therefore separated from Challenge-Based Learning on a different criterion again, treated in the companion review.
IV. What the evidence reports
A. Knowledge — mixed, and sensitive to how it is measured
Meta-analyses in medical education found problem-based students scoring slightly lower than traditionally taught students on multiple-choice basic-science examinations, and slightly better on clinical problem-solving, with small effect sizes on written clinical tests and moderate ones on performance ratings [1, p. 249]. A later meta-analysis found no effect on factual knowledge but a moderate effect favouring problem-based students on knowledge application [1, p. 249]. The qualification Hmelo-Silver draws is methodological: the studies support the approach “provided that the assessments measure knowledge in problem-solving contexts rather than in the context of multiple-choice examinations” [1, p. 250].
B. Reasoning strategies — the better-supported claim
Problem-based students were more likely to use hypothesis-driven reasoning and, in two studies, transferred that strategy to unrelated problems, producing more coherent explanations than students without the experience [1, p. 253]. They were also more likely to treat problem finding — deciding what the problem actually is — as a step when given a novel ill-structured problem [1, p. 253]. Their explanations were more elaborated than those of traditionally taught students, though also more error-prone; the reading offered is that “a well-elaborated knowledge structure that contains some errors” beats “little elaborated knowledge that cannot be applied” [1, p. 250].
C. Self-directed learning — real but uneven
Students in problem-based curricula were more likely to select their own learning resources rather than work from faculty-set assignments, and over four years relied progressively less on external scaffolds such as tests [1, pp. 255–256]. But becoming self-directed “is not a given”: learners with weak self-regulation focused on fact acquisition and struggled to adapt, and case studies show some students reverting to prior strategies rather than developing new ones [1, pp. 255–257].
D. Motivation — positive, undifferentiated, and unstable
A meta-analysis of 139 subsamples (83 of them problem-based) found a small to moderate pooled effect of d = 0.498 (95% CI [0.354, 0.641]) of problem-driven methods on motivation relative to lecture-based teaching, with high heterogeneity (I² = 75.12% between clusters) [5]. Crucially, problem-based, project-based and case-based learning did not differ (F(2, 285) = 0.61, p = .546) [5]. Effects were larger at course level than curriculum level, and larger in healthcare and STEM than elsewhere; the effect on students’ reasons for studying was trivial (d < 0.20) [5].
V. Scope of adoption
The literature is large — larger than that on any other active-learning approach [4] — and it is still a medical literature. Of the twenty most-cited PBL scholars, all but a few are associated with medical education, and the most frequent keywords are “medical education” (4,404 occurrences), “curriculum”, “teaching”, “medical students” and “clinical competence” [4]. Adoption is documented in nursing, dental, management, pharmaceutical, engineering, science and architectural education [4], while attention remains comparatively weak in business management, teacher training, social work, counselling and the social sciences [4]. The research front has been shifting toward self-directed learning, satisfaction, self-efficacy, critical thinking and cooperation [4].
Transfer is not automatic. A design-based study of an attempt to introduce the method into the Sri Lankan public university system reported that several of its basic assumptions did not hold: that team members share a common language, that students accept one another as equal peers, and that facilities for small-group work are available [8]. The authors add that in a high-power-distance setting, team leaders were expected to take responsibility for group work rather than the group owning it, and that changes confined to a single course unit could not succeed [8].
VI. Limitations
The evidence base does not cover the claims. Of the five goals, “there is considerable research on the first 3 … but little on the last 2” [1, p. 235] — that is, collaboration and intrinsic motivation, two of the method’s most-advertised benefits, are its least-evidenced. On collaboration specifically, “there is not yet evidence that supports the hypothesis that PBL helps students become better collaborators (broadly defined)” [1, p. 259], and not all groups collaborate well [1, p. 259].
The population studied is narrow and unusually able. “Minimal research has been conducted outside medical and gifted education” [1, p. 235]; there is little work with K-12 populations, and medical students are “a fairly select group” studying in well-established, whole-curriculum implementations that a single experimental course cannot reproduce [1, pp. 259–260].
The designs are mostly not experimental. Much of the research uses case-study, pre–post or quasi-experimental designs rather than controlled experiments [1, p. 260]. The motivation meta-analysis found publication bias (Egger’s test, t(137) = 4.19, p < .001) and calls for more randomised trials, noting that small quasi-experimental studies can overstate the effect of educational interventions [5].
Fidelity across the field is not assured. Critics quoted in the bibliometric review assert that some cross-disciplinary adaptations are “poorly designed and not carried out according to the core PBL model”, which complicates any attempt to synthesise findings within or across disciplines [4].
Portability is an open question, not a solved one. It “would be naïve to believe that the medical school model of PBL could be imported into other settings without considering how to adapt it to the local context, goals, and developmental level of learners” [1, p. 260] — a caution the Sri Lankan case makes concrete [8]. Self-directed learning may be particularly difficult for younger learners, and how to scaffold it for them “remains an empirical question” [1, p. 260].
References
[1] C. E. Hmelo-Silver, “Problem-Based Learning: What and How Do Students Learn?,” Educational Psychology Review, vol. 16, no. 3, pp. 235–266, Sep. 2004.
[2] J. R. Savery, “Overview of Problem-based Learning: Definitions and Distinctions,” Interdisciplinary Journal of Problem-Based Learning, vol. 1, no. 1, pp. 9–20, 2006, doi: 10.7771/1541-5015.1002. Quoted from the published abstract; the full text was not obtained.
[3] H. S. Barrows and R. M. Tamblyn, Problem-Based Learning: An Approach to Medical Education. New York, NY, USA: Springer, 1980. Cited for attribution only; not obtained for this review.
[4] P. Hallinger, “Tracking the Evolution of the Knowledge Base on Problem-based Learning: A Bibliometric Review, 1972–2019,” Interdisciplinary Journal of Problem-based Learning, vol. 15, no. 1, 2021, doi: 10.14434/ijpbl.v15i1.28984.
[5] L. Wijnia, G. Noordzij, L. R. Arends, R. M. J. P. Rikers, and S. M. M. Loyens, “The Effects of Problem-Based, Project-Based, and Case-Based Learning on Students’ Motivation: a Meta-Analysis,” Educational Psychology Review, vol. 36, no. 1, Art. 29, 2024, doi: 10.1007/s10648-024-09864-3.
[6] J. W. Thomas, “A Review of Research on Project-Based Learning,” The Autodesk Foundation, San Rafael, CA, USA, Research Review, Mar. 2000.
[7] B. Condliffe, J. Quint, M. G. Visher, M. R. Bangser, S. Drohojowska, L. Saco, and E. Nelson, “Project-Based Learning: A Literature Review,” MDRC, New York, NY, USA, Working Paper, Oct. 2017.
[8] D. Macan Markar, A. P. Madurapperuma, and J. Maroulis, “Problem-based Learning – is it right for Sri Lanka?,” in Proc. Asia Pacific Educational Research Association (APERA) Conf., Hong Kong, Nov. 2006.