Different early childhood classrooms all over the world have the same kind of scene. Four-year-olds are sitting around a table covered in blocks and quietly competing by putting the blocks in order by shape. Every once in a while, one of them will ask a question that seems to come out of nowhere. There is a teacher nearby who doesn’t say anything but nudges—”What do you think would happen if you put the bigger ones at the bottom?”—and then steps back. There’s nothing about it that looks like school. But more and more studies show that it may work better than anything else on the market.
A new study from the University of Cambridge makes a strong case that guided play, which is somewhere between unstructured free play and strict direct instruction, can be just as good as regular classroom teaching for kids ages three to eight, and in some ways be even better. The study used information from 39 studies that were done over more than 40 years and followed the progress of about 3,800 kids in reading, math, social skills, and something called “executive functions,” which is the ability to plan, concentrate, and switch between tasks. To put it simply, the results were better than what many teachers might have thought they would be.
The study doesn’t say that kids should be free to do whatever they want during guided play. It involves using open-ended questions and fun activities to guide kids slowly toward a certain learning goal while still letting them explore on their own. An adult might make a game that requires counting or a story-telling activity based on early reading ideas, but the child is the one who explores. It’s structured in a way that makes it feel free but also clear enough to be important.
One interesting thing about the math results is how big the effect sizes were. The difference between guided play and direct instruction in shape knowledge was 0.63, which is a medium-to-large statistical gain based on Hedge’s g scale that was used throughout the study. The number was 0.24 for early math skills in general. These aren’t huge numbers, but they’re important in the field of educational research. The most interesting thing about the headline is that there was no area where guided play did significantly worse than traditional instruction.
The researchers give a few reasons for why math in particular seems to work well with fun methods. This one makes sense: young kids have a hard time with abstract ideas like numbers, and putting those ideas into a physical, creative activity helps them see what’s really going on. Some people think that the logical, step-by-step nature of math problems is a lot like the gentle prompting style that guided play uses. There’s also the matter of motivation. Kids who are playing tend to stick with something longer, and persistence is important when you’re trying to really understand something hard.
One of the study’s co-authors, Dr. Elizabeth Byrne, said it in a way that sounds real rather than abstract. She said that play-based learning is like a spectrum, with free play on one end and direct instruction on the other. Guided play is somewhere in the middle. This way of putting things is helpful because it keeps us from seeing these approaches as opposites. The real question isn’t play vs. instruction. It depends on what kind of structure works best for each child at that moment.
When I read this research, I get the feeling that guided play for older kids hasn’t been taken as seriously as it should be in the world of education. It happens a lot in preschools. It happens less often as kids get older and start primary school. That’s when formal lessons start, and “playing” starts to feel like something that needs to be explained. There are real questions about when structure is needed, so that feeling isn’t completely wrong. But the Cambridge results suggest that the move away from play-based learning may be happening sooner than the evidence supports.
Right now, the researchers want more research to be done on how guided play affects learning: what parts are most important, what subjects respond best, and how teachers can be helped to do it well. Not all of that work has been done yet. The base is there, though. Things we thought the sandbox could only do have been done by it.
