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 Project-based learning. The review date below matters, because the literature moves.
A definitional review. Citations follow IEEE style; see References.
Abstract
Project-Based Learning is an instructional model that organises learning around sustained projects culminating in a product or performance. It shares the acronym PBL with Problem-Based Learning, a separate pedagogy with a different origin, unit of work and assessment logic; this article covers the project sense only, and the companion review covers the other. Set out here are the criteria the literature uses to screen a project in, the lineage, the features that separate it from adjacent inquiry pedagogies, what the empirical record reports, and what it does not establish.
I. Definition
Project-Based Learning “is a model that organizes learning around projects” [1]. Synthesising the definitions then in circulation, Thomas characterised projects as
complex tasks, based on challenging questions or problems, that involve students in design, problem-solving, decision making, or investigative activities; give students the opportunity to work relatively autonomously over extended periods of time; and culminate in realistic products or presentations [1].
Because that description admits almost any classroom activity called a project, the same review offers five criteria designed to answer a narrower question — “what must a project have in order to be considered an instance of PBL?” [1]:
- Centrality — the project is the curriculum, not an illustration bolted onto instruction delivered by other means. “Application” projects and “enrichment” projects are excluded [1].
- Driving question — the project is built on a question or ill-defined problem that drives learners to encounter and struggle with the central concepts of a discipline [1].
- Constructive investigation — the central activity must transform and construct knowledge. If it can be completed with already-learned skills, “the project is an exercise, not a PBL project” [1].
- Autonomy — projects are student-driven to a significant degree, and do not “end up at a predetermined outcome or take predetermined paths” [1].
- Realism — projects are “realistic, not school-like”, in topic, task, role, context, collaborators, product, audience or judging criteria [1].
These are criteria, not a definition, and the distinction matters: the field has never converged on a single model. A later review found “little consensus among developers of PBL design principles … about how PBL fits in with other instructional methods, how long a PBL unit should last, the roles of student choice and collaborative learning, and how learning should be assessed” [2].
II. Origins
The lineage runs through progressive education. Reviews trace the approach to John Dewey’s philosophy and to William Heard Kilpatrick’s project method of 1918, “cited as the first formalization of a PBL model”, whose key idea was “an activity undertaken by students that really interested them” [2]; historians have found evidence of learning through projects in the seventeenth and eighteenth centuries, so 1918 is a formalisation rather than an invention [2]. A parallel account adds the vocational educator David Snedden, who advocated practical projects for learning by doing [3].
The modern research programme is much younger. Writing in 2000, Thomas observed that “all of the research on Project-Based Learning has taken place in the past ten years and most of it in just the last few years” [1], and identified three traditions feeding it: Outward Bound wilderness expeditions, by way of Expeditionary Learning; postsecondary problem-based learning; and university research in cognition — motivation, expert–novice differences, situated cognition and technology as a cognitive tool [1]. The contemporary school-reform strand is usually dated to Blumenfeld and colleagues’ work in science education in the 1990s [5], [8].
III. What distinguishes it
A. From Problem-Based Learning — the acronym collision
“PBL” denotes Project-Based Learning in the school-reform literature and Problem-Based Learning in medical and health-professions education. They are not variants of one term. Problem-Based Learning was “developed in the late 1960s to reform medical education at McMaster University (Canada)” and refined at Maastricht from 1974; Project-Based Learning descends from Kilpatrick’s project method and the 1990s science-education work [5]. Readers arriving from either literature should expand the acronym on first use; the companion review covers the problem-based sense.
A meta-analysis that coded both separately sets out the operational differences [5]:
| Project-Based | Problem-Based | |
|---|---|---|
| Unit of work | a project run over weeks or months | one ill-structured problem, usually two sessions |
| Product | always culminates in a tangible product or presentation | optional, in the reporting phase |
| Role of the problem | provides the context for the driving question | starts the learning process and activates prior knowledge |
| Student control | high — the process is student-driven | medium–high — the tutor checks learning goals against intended objectives |
The consequence for assessment follows from the first two rows: project-based work is judged on a sustained artefact, problem-based work on the reasoning a group performs on a case. Read from the problem-based side, the same comparison is drawn on the type and role of the problem, the process, the teacher’s role, the mode of collaboration and the tools — with project-based science treated as a sibling approach in which the teacher introduces relevant content before and during inquiry, rather than withholding it [6, p. 238].
The boundary is nevertheless contested, and honest reviews say so. Thomas held that problem-based learning studies “have all of the defining features of PBL (centrality, driving question, constructive investigation, autonomy, and realism)” and included them in his review [1]. Others argue the distinction is load-bearing, “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” [2, citing 7]. The MDRC review excluded the “close cousins” — problem-based and inquiry-based learning among them — precisely because the definitional overlap made pooled findings unsafe [2]. A second collision is worth flagging: in the motivation meta-analysis, CBL abbreviates case-based learning, not challenge-based learning [5].
B. From Challenge-Based Learning
Project-Based Learning’s realism criterion concerns the authenticity of the work — its topic, context, collaborators, audience and judging criteria may all be drawn from outside the school, and Thomas endorses “real-life challenges where the focus is on authentic (not simulated) problems or questions and where solutions have the potential to be implemented” [1]. But nothing in the criteria requires a party outside the institution to own the problem or judge the result; authenticity is designed for, not procured. That is the axis on which Challenge-Based Learning is separated in the companion review, and it is why a project can be entirely realistic and still not be a challenge.
IV. What the evidence reports
The comparison condition throughout is “teacher-centered/lecture-based learning” [5] or “traditional, teacher-directed instruction” [1] — rarely a single well-specified model on either side, which is the source of most of the interpretive difficulty below.
A. Cognitive outcomes
The strongest single result available is a cluster-randomised trial with 48 second-grade teachers in low-socioeconomic-status U.S. districts. Controlling for baseline and student characteristics, the project-based group outscored the comparison group on social studies by ES = 0.482 (p < .001) — above the What Works Clearinghouse threshold of 0.25 for “substantively important” — and on informational reading by ES = 0.182 (p = .083), below that threshold. Effects on informational writing (ES = −0.047) and motivation (ES = 0.135) were not statistically significant [3]. The authors translate the social-studies result as roughly five to six months of additional learning [3].
An earlier randomised trial of a problem-based economics curriculum aligned with project-based design principles reported effects of 0.32 on economic literacy and 0.27 on problem-solving in economics, though the study suffered substantial teacher attrition [2].
B. Fidelity, not the label, carries the effect
In the same trial, consistency with the unit session plans was associated with higher scores on every measure — social studies 0.270, reading 0.583, writing 0.239, motivation 0.287 — all above the 0.25 threshold, including the two outcomes on which the intervention itself showed no effect [3]. Classrooms with the lowest fidelity had eliminated session components, substituted whole-class teaching for the specified small-group work, and left no time for review and reflection [3].
C. Motivation
A meta-analysis of 139 subsamples from 132 reports (37 of them project-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 substantial heterogeneity (I² = 75.12% between clusters) [5]. Two qualifications matter. First, no statistically significant difference was found between project-based, problem-based and case-based learning (F(2, 285) = 0.61, p = .546) — on this outcome the choice of method was not what mattered [5]. Second, effects were larger at course level than at curriculum level, and larger in healthcare and STEM than elsewhere [5]. The effect on students’ reasons for studying — intrinsic versus extrinsic motivation — was trivial (d < 0.20) [5].
D. Implementation is the recurring constraint
Enactment problems reported for teachers new to the approach cluster under six headings: time, classroom management, control of information flow, support of student learning, technology use and assessment [1]. A multiple-case study of teachers already active in project work found that collaboration, artefacts, technological tools, problem-centredness and scientific practices were readily promoted, but that driving questions, learning goals set by students, students’ own questions, the integrity of the project activities, and using the project as the means to learn central content were harder to implement [4] — that is, precisely the criteria that distinguish project-based learning from ordinary project work proved the ones most often lost.
V. Scope of adoption
Project work enters classrooms by three routes: externally developed curricula, teacher-initiated design, and whole-school reform [2]. Research is unevenly distributed across them. Most implementation research concerns externally developed, “packaged” curricula administered by teachers with limited prior experience, while “teacher-initiated PBL is likely the most common way that students are exposed to a PBL approach, but it is the least well understood” [2] — a gap first flagged in 2000 [1] and still open seventeen years later [2]. Studies concentrate in science and social studies; mathematics and English/language arts are comparatively unstudied [2].
Practitioner-facing claims run ahead of this. A funder-published summary of four commissioned studies states that “the evidence is clear”, reporting that a project-based Advanced Placement programme raised the probability of a credit-qualifying exam score by about eight percentage points in the first year of implementation and ten in the second [9]. The trials are real; the framing is the funder’s.
VI. Limitations
The absence of a shared model is the primary limitation, and it is upstream of every effect size. Because studies “do not share common design principles … it is difficult to use the existing body of research to draw conclusions about PBL’s effectiveness” [2]. Reviews are aggregating interventions that agree on a name and little else.
Much of the outcome literature cannot support causal claims. Except for a handful of trials, “most of the quantitative outcome studies … must be considered descriptive because concerns about internal validity preclude causal inferences” [2]. Selection bias is the specific worry: the dispositions that lead a school, teacher or student to opt into an innovative approach may themselves produce the outcomes attributed to it [2].
The outcomes the approach claims are the ones hardest to measure. Intrapersonal and interpersonal competencies remain under-assessed for want of valid, reliable, scalable instruments, so studies of cognitive outcomes “continue to far outnumber” them [2] — a gap identified in 2000 and unclosed [1], [2]. The motivation literature carries its own: publication bias was detected (Egger’s test, t(137) = 4.19, p < .001), and trim-and-fill adjusted the pooled estimate upward from 0.50 to 0.766, which the authors treat as grounds for caution rather than encouragement [5].
Effects reported for whole-school programmes may not be attributable to project work at all. Thomas noted that gains in Expeditionary Learning and Co-nect schools could owe to portfolios, block scheduling or technology, and that the findings were drawn from those programmes’ own publications and “undoubtedly selected for their salience and positive direction” [1]. The settled verdict of the most careful review remains the cautious one: the evidence is “promising but not proven” [2].
References
[1] J. W. Thomas, “A Review of Research on Project-Based Learning,” The Autodesk Foundation, San Rafael, CA, USA, Research Review, Mar. 2000.
[2] 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.
[3] N. K. Duke, A.-L. Halvorsen, S. L. Strachan, J. Kim, and S. Konstantopoulos, “Putting PjBL to the Test: The Impact of Project-Based Learning on Second Graders’ Social Studies and Literacy Learning and Motivation in Low-SES School Settings,” American Educational Research Journal, 2021, doi: 10.3102/0002831220929638.
[4] A. Markula and M. Aksela, “The key characteristics of project-based learning: how teachers implement projects in K-12 science education,” Disciplinary and Interdisciplinary Science Education Research, vol. 4, no. 2, 2022, doi: 10.1186/s43031-021-00042-x.
[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] 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.
[7] 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. Full text not obtained; cited here only as reported in [2].
[8] P. C. Blumenfeld, E. Soloway, R. W. Marx, J. S. Krajcik, M. Guzdial, and A. Palincsar, “Motivating Project-Based Learning: Sustaining the Doing, Supporting the Learning,” Educational Psychologist, vol. 26, no. 3–4, pp. 369–398, 1991, doi: 10.1080/00461520.1991.9653139. Cited for attribution only; full text not read for this review.
[9] “The Evidence Is Clear: Rigorous Project-Based Learning Is an Effective Lever for Student Success,” Lucas Education Research, research summary, n.d. Funder-published summary of four commissioned studies, not an independent review.