John Dewey and Progressive Education: Learning by Doing, Democracy, and the Research Evidence

John Dewey's progressive education, first tested at the University of Chicago Laboratory School he founded in 1896, reimagined the classroom as a miniature community in which children learn through purposeful activity, shared projects, and democratic life rather than recitation and drill. This review traces the tradition from its Chicago origins through its twentieth-century rise and retreat, then examines the modern research evidence bearing on its central claims. That evidence is mixed in instructive ways. Unguided discovery learning consistently underperforms explicit teaching, yet well-scaffolded inquiry and project-based learning — far closer to what Dewey actually proposed — show meaningful positive effects in meta-analyses and in a rigorous randomized trial in early elementary classrooms. Research on social-emotional learning and open classroom discussion supports Dewey's insistence that schools are small societies with civic consequences, and early-childhood studies favor child-centered over didactic instruction chiefly on motivational grounds. The review closes with practical implications for families and a candid account of what remains unproven.

In January 1896, a small elementary school opened on the South Side of Chicago with sixteen children and an outsized ambition. Its founder, John Dewey (1859–1952), born in Burlington, Vermont, had arrived at the University of Chicago two years earlier to lead a department spanning philosophy, psychology, and pedagogy, and he conceived the school as a working laboratory: a place where hypotheses about how children learn could be tried, observed, and revised, much as a chemist tests ideas at the bench (Hildebrand, 2024; Tanner, 1997). His wife, Alice Chipman Dewey, served as the school's principal. The curriculum was organized not around recitation in separate subjects but around what Dewey called occupations — cooking, sewing and textile work, carpentry, gardening — treated not as job training but as socially meaningful activities through which children could be led into history, science, number, and language. In The School and Society (1899), the published lectures he gave to the school's parents, Dewey argued that when a school takes children's purposeful activity as its starting point, it becomes, in his words, "a miniature community, an embryonic society" (Dewey, 1899).

Two years before those lectures, Dewey had compressed his outlook into My Pedagogic Creed (1897), a five-part manifesto in which every paragraph begins with the words "I believe." Its core claims are that education is fundamentally a social process; that the school is a form of community life rather than a preparation for some future living; and that the child's own powers and interests supply the raw material with which the educator must work. It is worth stressing at the outset what Dewey did not say. He was as impatient with sentimental child-worship as with rote drill: interests, he argued, are signs of growing capacity to be interpreted and directed by the teacher, not whims to be indulged. Late in life, in Experience and Education (1938), he rebuked the "new" schools founded in his name as sharply as the traditional ones, insisting that experience by itself educates nothing — experiences can be educative or mis-educative, and it is precisely the teacher's responsibility to organize subject matter progressively out of children's experience rather than abandon organization altogether.

Dewey resigned from Chicago in 1904 after conflicts over the school's administration — a story long blamed on University President William Rainey Harper, though archival work by Knoll (2015) locates the decisive friction in faculty opposition to Alice Dewey's contested principalship — and spent the rest of his career at Columbia University. There he produced Democracy and Education (1916), his most systematic treatise, which grounds schooling in a striking claim: democracy is not merely a form of government but a mode of associated living, and schools serve democracy only insofar as they are themselves communities of shared inquiry in which children learn to communicate across differences. The broader movement he inspired institutionalized itself in the Progressive Education Association, founded in 1919, and washed through American schooling in waves that the historian Lawrence Cremin (1961) chronicled in his classic account of progressivism from 1876 to 1957. Cremin's central lesson bears on everything that follows: "progressive education" was never one thing. It encompassed child-centered romantics, administrative efficiency reformers, and social reconstructionists, and much that traveled under the banner would have earned Dewey's rebuke. Evaluating "Dewey's progressive education" against research therefore requires care about which version is on the table.

Inside the Laboratory School: a rich record, not a dataset

The primary evidence about the Laboratory School itself is documentary. The fullest account, The Dewey School (Mayhew & Edwards, 1936), was written by two of its teachers, with an introduction by Dewey, and preserves in detail how seven years of curriculum were built around occupations, group work, and the reconstruction of human history through children's own activity. Tanner (1997) mined the school's teacher reports for lessons still usable today, documenting practices — integrated curriculum, sustained teacher collaboration and weekly staff study, discipline handled as a problem of community rather than obedience — that anticipate much of what contemporary reform literature recommends. But honesty requires stating plainly what this record is not: the school ran for less than a decade under Dewey, enrolled a small and largely advantaged population, and was never evaluated against any comparison group. Its afterlife has been as much mythology as history. Fallace and Fantozzi (2017), reviewing 120 years of writing about the school, identify three recurring narratives — the school as triumph, as tragedy, and as misunderstood — and show that these are largely historiographical constructions; Dewey himself endorsed only the last. The Laboratory School is best treated as the movement's founding demonstration and thought experiment, not as outcome evidence. For that, we must turn to the modern literatures that test its component claims.

The school's institutional legacy, by contrast, is easy to trace. The experiment Dewey began with sixteen pupils survives today as the University of Chicago Laboratory Schools, and it inaugurated the American laboratory-school movement that joined teacher preparation to educational research. Dewey's ideas also seeded other durable institutions of early-childhood progressivism. Lucy Sprague Mitchell, working consciously in the Deweyan stream, founded the Bureau of Educational Experiments in New York in 1916 to study the environments in which young children learn to their full potential; her bureau's early conclusion — that children's emotional lives are inseparable from their learning and motivation — anticipated decades of developmental science, and the organization she built became Bank Street College of Education and its School for Children, whose developmental-interaction approach still names Dewey among its foundations. The Progressive Education Network, successor to the 1919 association, continues to link progressive schools and educators across the United States. Progressivism as an organized movement collapsed at mid-century, but its institutional carriers never entirely disappeared — which is one reason so many of its practices were available for modern research to test.

Learning by doing under the microscope: the guidance debate

The most direct scientific challenge to Deweyan pedagogy is the instructional-guidance debate. In a widely cited polemic, Kirschner, Sweller, and Clark (2006) argued that "minimally guided" approaches — a category in which they grouped discovery, problem-based, inquiry, experiential, and constructivist teaching — fail because they ignore human cognitive architecture: novices lack the organized background knowledge to search a problem space productively, and unguided exploration overloads working memory while leaving little behind in long-term memory. The empirical record gives real support to part of this critique. Alfieri, Brooks, Aldrich, and Tenenbaum (2011), meta-analyzing 164 studies, found that unassisted discovery produced worse outcomes than explicit instruction across 580 comparisons. If Dewey's legacy were the caricature — turn children loose and learning will happen — the verdict would be in, and it would be negative.

But the same meta-analysis contained a second finding that reverses the moral: across 360 comparisons, enhanced discovery — exploration supported by guidance, feedback, scaffolding, and worked examples — outperformed other forms of instruction, including explicit teaching. Hmelo-Silver, Duncan, and Chinn (2007), replying to Kirschner and colleagues, made the corresponding conceptual point: genuine problem-based and inquiry learning are not minimally guided at all but heavily scaffolded, with teachers structuring tasks, modeling reasoning, and progressively transferring responsibility. Lazonder and Harmsen (2016) then quantified this in a meta-analysis of 72 studies of inquiry learning, finding that guidance produced substantial benefits for learning activities (d = 0.66), task performance (d = 0.71), and learning outcomes (d = 0.50), with structuring assignments and access to domain information among the most effective supports. The consensus emerging from this literature is that the operative variable is not "doing versus listening" but the quality and placement of guidance within activity.

What is striking, reading this debate a century downstream, is how much of it Dewey adjudicated in advance. Experience and Education (1938) is, in effect, a rebuke of minimally guided instruction written by the tradition's own founder: freedom of activity without intelligent direction, Dewey wrote there, is as mis-educative as external imposition, and the harder task — the one the new schools too often shirked — is selecting and organizing experiences so that each leads fruitfully into the next. The modern evidence does not vindicate everything done in Dewey's name. It does, rather precisely, vindicate the position Dewey actually held: experiential learning works when it is deliberately scaffolded by a knowledgeable teacher, and fails when it is not.

From the project to project-based learning: the modern trials

The pedagogical vehicle most identified with progressive education is the project — an extended, purposeful undertaking through which academic content is learned in use. Popularized in the 1910s by Dewey's colleague William Heard Kilpatrick (in a form Dewey thought too child-driven), the project survives today as project-based learning (PBL), whose research base has matured considerably. A methodical literature review commissioned by MDRC (Condliffe et al., 2017) traced PBL's design principles directly to Dewey, and concluded that the approach is promising but that the evidence base was still limited: definitions vary across studies, implementation is demanding for teachers, and many evaluations have weak designs — a caution worth keeping in view whenever enthusiasts cite the literature wholesale.

Within those limits, the quantitative signal is positive. Chen and Yang (2019), meta-analyzing studies from primary through tertiary education, estimated a mean weighted effect of d = 0.71 for PBL over traditional instruction on academic achievement — a medium-to-large effect — with the effect moderated by subject area, instructional hours, and technology support, though not by educational stage. Because meta-analyses inherit the weaknesses of their ingredient studies, the more telling evidence comes from a large randomized experiment in exactly the age range early educators care about. Duke, Halvorsen, Strachan, Kim, and Konstantopoulos (2021) randomly assigned 48 second-grade teachers in low-income school districts, in matched within-school pairs, to teach four project-based social studies units or to continue business as usual. Children in the project-based classrooms showed significantly higher growth in social studies knowledge and in informational reading than controls — a rare experimental demonstration that a Dewey-descended method can raise achievement for young, disadvantaged learners. The same trial also illustrates the honest limits of the evidence: no advantages appeared for writing or for motivation, a reminder that PBL is not a universal solvent even where it works.

An embryonic society: social, emotional, and civic outcomes

Dewey's deepest claim was never about test scores. It was that the school is a society in miniature, and that what children practice socially — cooperation, communication, shared decision-making — is curriculum, with consequences for the democratic communities they will join (Dewey, 1897; Dewey, 1916). Two modern literatures speak to this claim. The first is research on social and emotional learning. Durlak, Weissberg, Dymnicki, Taylor, and Schellinger (2011), in a meta-analysis of 213 universal school-based programs enrolling more than 270,000 students, found that deliberately teaching social and emotional skills improved not only behavior and attitudes but academic achievement, by an average of 11 percentile points. Modern SEL programs are more structured and skill-focused than anything in Dewey's writing, so the finding should not be over-attributed to him; what it confirms is his underlying premise that social development is not a distraction from academic schooling but a contributor to it. The meta-analysis's fine print is also instructive: programs worked best when skill practice was sequenced, active, focused, and explicit — the "active" ingredient echoing Dewey's learning by doing, and the "sequenced" and "explicit" ingredients echoing, once again, his insistence that experience must be deliberately organized to educate.

The second literature concerns the democratic classroom itself. Campbell (2008), analyzing data from the large international CIVED study of civic education, found that an open classroom climate — one in which students freely discuss contested social and political questions and feel their voice is welcome — predicted greater civic knowledge and appreciation of legitimate political conflict, along with stronger intentions to become informed voters, after controlling for individual, classroom, school, and district characteristics. Notably, the benefit was largest for students of low socioeconomic status, for whom open classroom discussion partially compensated for civic disadvantages at home. These are observational data, not an experiment, and self-selection into discussion-rich classrooms cannot be fully excluded. Still, it is difficult to imagine a finding more congruent with Democracy and Education: the classroom that functions as a small deliberative community appears to be a working apprenticeship in citizenship, and most valuable to the children with fewest alternatives.

The Eight-Year Study: progressive education's grand evaluation

Progressive education also produced one of the earliest large-scale evaluations in American educational history. Between the early 1930s and 1941, the Progressive Education Association's Commission on the Relation of School and College persuaded some thirty secondary schools — a mix of public, independent, and university-affiliated institutions — to redesign their curricula along progressive lines, with integrated and interdisciplinary programming, expanded arts, collaborative teaching, and community involvement, while more than 200 colleges agreed to admit their graduates without conventional entrance requirements. The graduates were then followed through college in comparison with matched peers from conventional schools. As reported by the study's director (Aikin, 1942), the experimental-school graduates did as well as their matched counterparts academically and were more engaged in artistic, political, and social life; graduates of the schools that had departed most radically from the traditional pattern showed the strongest college records of all.

The Eight-Year Study deserves both its fame and its footnotes. Matching is not randomization; the participating schools served largely advantaged students and received consulting, funding, and publicity no ordinary school could expect; and its reforms did not survive the postwar decade, as wartime staffing pressures, increased teacher workload, rising college selectivity, and the conservative turn Cremin (1961) documents rolled progressive practice back — the Progressive Education Association itself dissolved in 1955. Read carefully, the study does not prove progressive schooling superior. It established something more modest and still important: students educated outside the conventional academic sequence suffered no detectable academic penalty at college, and showed broader engagement — an early, imperfect anticipation of the pattern modern research keeps finding for well-implemented alternatives to traditional instruction.

Progressive practice in the earliest years

For families of preschoolers, the most pointed question is what Dewey's ideas imply for children of four, five, and six — the ages his Laboratory School's youngest groups served with play, practical activity, and social occupations rather than formal lessons. Here the child-centered tradition and the modern evidence align fairly well, with caveats. Stipek, Feiler, Daniels, and Milburn (1995), comparing 227 four-to-six-year-olds in didactic, basic-skills programs against peers in child-centered programs, found that didactic instruction bought a modest advantage on a letters-and-reading test (and none in mathematics) at a measurable motivational price: children in didactic programs rated their own abilities lower, expected less success, depended more on adults for approval, showed less pride in accomplishment, and reported more worry about school. Marcon (2002) found that graduates of child-initiated preschool classrooms earned better grades by the end of elementary school than peers from academically directed preschools, whose performance declined after fourth grade — though as a non-randomized study its causal interpretation has been debated. Neither study evaluated a "Dewey program" as such, and the didactic-versus-child-centered contrast is coarser than Dewey's own position, which called for adult-guided purposeful activity rather than either drill or free play. But the pattern — early didactic pressure yields small, fading academic gains alongside motivational costs — is consistent across this literature, and it is the strongest empirical argument that the early years, at minimum, should look more like Dewey's kindergarten than like a scaled-down academy.

Practical implications

Several implications follow for parents and educators weighing progressive, Dewey-inflected programs. First, discount both the romance and the caricature: the research does not support unstructured discovery, and neither did Dewey. The question to ask of any experiential program is not whether children are active but how teachers guide the activity — what scaffolds, feedback, and organized subject matter surround the doing (Alfieri et al., 2011; Lazonder & Harmsen, 2016). Second, projects are a genuine asset when they are well designed: the best current evidence, including a randomized trial with second graders in low-income schools, shows that teacher-structured, standards-connected project units can outperform conventional instruction in content knowledge and informational reading (Duke et al., 2021), though families should expect thoughtful unit design and trained teachers, not projects as decoration. Third, take the classroom-as-community seriously as an academic matter, not a soft extra: social-emotional skill-building measurably supports achievement (Durlak et al., 2011), and classrooms where children's voices are welcomed in real discussion build civic capacity, especially for children with fewer advantages (Campbell, 2008). Fourth, for the preschool and kindergarten years specifically, prefer programs organized around purposeful, adult-guided activity over didactic drill; the academic advantages of early didactic instruction appear small and transient, while the motivational costs are consistently documented (Stipek et al., 1995; Marcon, 2002). Finally, apply Dewey's own test to any school claiming his mantle: ask not whether it has abandoned the old order but what it has deliberately organized in its place.

A century and a quarter after sixteen children gathered in Dewey's experimental school, the research literature has done something unexpected to John Dewey's legacy: it has discriminated within it. The versions of progressivism he himself criticized — child-whim romanticism, activity without direction — find little support in the evidence and considerable evidence against them. The version he actually argued for — purposeful activity deliberately scaffolded by knowledgeable teachers, projects that carry real subject matter, a classroom that functions as a small society practicing democratic life — is supported by guided-inquiry meta-analyses, an emerging body of project-based learning trials, social-emotional learning research, and civic-education findings, with honest gaps remaining around writing, motivation, and long-term outcomes. Dewey believed the Laboratory School was chiefly misunderstood, and the historiography suggests he was right (Fallace & Fantozzi, 2017). The research of the past three decades may be read as a slow correction of that misunderstanding: not a vindication of everything called progressive, but of the harder, more disciplined idea at its center — that children learn best when their own purposeful experience is taken seriously enough to be organized well.

References

1. Aikin, W. M. (1942). The story of the Eight-Year Study: With conclusions and recommendations. New York: Harper & Brothers. https://archive.org/details/storyoftheeighty009637mbp
2. Alfieri, L., Brooks, P. J., Aldrich, N. J., & Tenenbaum, H. R. (2011). Does discovery-based instruction enhance learning? Journal of Educational Psychology, 103(1), 1–18. https://eric.ed.gov/?id=EJ933606
3. Campbell, D. E. (2008). Voice in the classroom: How an open classroom climate fosters political engagement among adolescents. Political Behavior, 30(4), 437–454. https://doi.org/10.1007/s11109-008-9063-z
4. Chen, C.-H., & Yang, Y.-C. (2019). Revisiting the effects of project-based learning on students' academic achievement: A meta-analysis investigating moderators. Educational Research Review, 26, 71–81. https://www.sciencedirect.com/science/article/abs/pii/S1747938X19300211
5. Condliffe, B., Quint, J., Visher, M. G., Bangser, M. R., Drohojowska, S., Saco, L., & Nelson, E. (2017). Project-based learning: A literature review (Working paper). New York: MDRC. https://eric.ed.gov/?id=ED578933
6. Cremin, L. A. (1961). The transformation of the school: Progressivism in American education, 1876–1957. New York: Alfred A. Knopf. https://archive.org/details/transformationof0000crem
7. Dewey, J. (1897). My pedagogic creed. The School Journal, 54, 77–80. http://dewey.pragmatism.org/creed.htm
8. Dewey, J. (1899). The school and society. Chicago: University of Chicago Press. https://archive.org/details/schoolsociety00deweuoft
9. Dewey, J. (1916). Democracy and education: An introduction to the philosophy of education. New York: Macmillan. https://www.gutenberg.org/ebooks/852
10. Dewey, J. (1938). Experience and education. Kappa Delta Pi Lecture Series. New York: Macmillan. https://openlibrary.org/works/OL111355W/Experience_and_education
11. Duke, N. K., Halvorsen, A.-L., Strachan, S. L., Kim, J., & Konstantopoulos, S. (2021). 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, 58(1), 160–200. https://doi.org/10.3102/0002831220929638
12. Durlak, J. A., Weissberg, R. P., Dymnicki, A. B., Taylor, R. D., & Schellinger, K. B. (2011). The impact of enhancing students' social and emotional learning: A meta-analysis of school-based universal interventions. Child Development, 82(1), 405–432. https://doi.org/10.1111/j.1467-8624.2010.01564.x
13. Fallace, T., & Fantozzi, V. (2017). The Dewey School as triumph, tragedy, and misunderstood: Exploring the myths and historiography of the University of Chicago Laboratory School. Teachers College Record, 119(2), 1–32. https://doi.org/10.1177/016146811711900205
14. Hildebrand, D. (2024). John Dewey. In E. N. Zalta & U. Nodelman (Eds.), The Stanford Encyclopedia of Philosophy. Stanford, CA: Metaphysics Research Lab, Stanford University. https://plato.stanford.edu/entries/dewey/
15. Hmelo-Silver, C. E., Duncan, R. G., & Chinn, C. A. (2007). Scaffolding and achievement in problem-based and inquiry learning: A response to Kirschner, Sweller, and Clark (2006). Educational Psychologist, 42(2), 99–107. https://doi.org/10.1080/00461520701263368
16. Kirschner, P. A., Sweller, J., & Clark, R. E. (2006). Why minimal guidance during instruction does not work: An analysis of the failure of constructivist, discovery, problem-based, experiential, and inquiry-based teaching. Educational Psychologist, 41(2), 75–86. https://doi.org/10.1207/s15326985ep4102_1
17. Knoll, M. (2015). John Dewey as administrator: The inglorious end of the Laboratory School in Chicago. Journal of Curriculum Studies, 47(2), 203–252. https://doi.org/10.1080/00220272.2014.936045
18. Lazonder, A. W., & Harmsen, R. (2016). Meta-analysis of inquiry-based learning: Effects of guidance. Review of Educational Research, 86(3), 681–718. https://doi.org/10.3102/0034654315627366
19. Marcon, R. A. (2002). Moving up the grades: Relationship between preschool model and later school success. Early Childhood Research & Practice, 4(1). https://ecrp.illinois.edu/v4n1/marcon.html
20. Mayhew, K. C., & Edwards, A. C. (1936). The Dewey School: The Laboratory School of the University of Chicago, 1896–1903. New York: D. Appleton-Century. https://archive.org/details/deweyschoolthela008095mbp
21. Stipek, D., Feiler, R., Daniels, D., & Milburn, S. (1995). Effects of different instructional approaches on young children's achievement and motivation. Child Development, 66(1), 209–223. https://www.semanticscholar.org/paper/Effects-of-Different-Instructional-Approaches-on-Stipek-Feiler/dcafd3733fd043fd0524a648ddd44f0c01488005
22. Tanner, L. N. (1997). Dewey's Laboratory School: Lessons for today. New York: Teachers College Press. https://www.tcpress.com/dewey-s-laboratory-school-9780807736180