Research into alternative education almost always sparks debate. Not only because people tend to have strong opinions about Montessori, Dalton, or Freinet schools, but also because conducting solid research on them is surprisingly difficult. Yesterday, while going through my archive of studies I had once intended to blog about, I stumbled upon this Dutch study. Vivian Morssink-Santing and colleagues specifically examined the transition from Montessori and Dalton secondary education to higher education. And this very solid study once again illustrates those challenges beautifully.

So, to be clear: this is not a bad study. Quite the opposite. I want to discuss it precisely because it is a fairly strong study. At the same time, it also shows how cautious our conclusions sometimes need to be.

The researchers examined nearly 9,000 first-year higher-education students, a smaller subset of whom had attended Dalton or Montessori schools. They investigated two things: how well students felt their secondary education prepared them for higher education, and how strong their self-regulated learning skills were. Think of skills such as planning, working independently, motivation, and dealing with procrastination.

At first glance, that seems perfectly logical. Alternative education often places greater emphasis on autonomy, independence, and responsibility. You might therefore expect these students to transition more smoothly into higher education, where independent learning becomes increasingly important.

But this is exactly where the first problem with this kind of research begins: what exactly counts as “Montessori” or “Dalton”? Even the authors acknowledge that schools within the same movement can differ substantially from one another. Montessori’s ideas about secondary education were never fully developed in a concrete way. Dalton schools have traditionally been given considerable freedom in implementing Dalton principles. As a result, researchers are not really comparing one clearly defined method with another, but rather broad families of schools operating under the same label.

There is another complication as well. Many characteristics once considered typical of alternative education have gradually found their way into mainstream schools. Independent work, group work, reflection, and attention to self-regulation are no longer exclusive to Montessori or Dalton education. The researchers themselves point to the influence of the Dutch “Study House” reforms, which pushed mainstream schools toward more independent learning practices. This creates a second problem: the distinction between “alternative” and “mainstream” education has become less clear.

A third difficulty, one I have written about before, concerns selection effects. Parents do not choose alternative education randomly. Beliefs, socio-economic background, and expectations often play a role. That makes it particularly difficult to determine whether any observed differences are truly caused by the education itself or by the families choosing that type of schooling. The authors explicitly acknowledge that they lacked sufficient data on socio-economic status and prior school performance to fully control for this. Interestingly, some Montessori studies have tried to address this issue more cleverly by comparing students who did and did not obtain a place through school waiting lists.

The way these issues are measured also remains challenging. In this study, for example, the measures relied largely on self-reporting. Students had to indicate how good they believed they were at planning, discipline, or independent study. But people who think more deeply about self-regulation may also judge themselves more critically. Those who better understand what effective planning actually involves may simply become more aware of their own shortcomings. All of this makes interpretation difficult.

And then we arrive at perhaps the most interesting finding of the study: the effects ultimately turn out to be small. Very small, in fact. Dalton students rated the transition to higher education slightly more positively, but the explained variance remained limited. For self-regulated learning, the differences disappeared once the researchers included other factors in the statistical models.

That does not mean alternative education “doesn’t work” or, in this case, “doesn’t work better.” But it does suggest that education is probably more complex than a single label on a school gate. The real differences may lie more in individual teachers, school culture, or specific instructional choices than in the name of a pedagogical movement itself, something Margaret Brown already described back in 2012.

So perhaps the best conclusion is not that Montessori or Dalton are magical solutions. But neither should we conclude that they lack valuable practices. Instead, this study may simply remind us that good education is difficult to reduce to a single method.