My daughter, Cee, is almost two-and-a-half. Five mornings per week, she is cared for by an in-home childcare provider I’ll call Amanda. On a given day, Amanda and an assistant care for between four and eight kids ranging in age from one to six years old. The one-year-olds are just starting to walk. The six-year-old, who also attends half-day kindergarten, can read and write.

One of the reasons why we chose Amanda’s for childcare is that it allows Cee to interact with kids of different ages. When I first visited Amanda’s, the problem-solving and conflict resolution that I observed among the older kids impressed me. It was refreshing to watch, particularly after visiting chaotic childcare centers with rooms of as many as 10 two-year-olds. I imagined that Cee would learn so much from observing and playing with these more mature children.

When Cee began daycare last fall, we watched as two older girls took her under their wing, reading her stories and including her in their pretend play. Suddenly Cee was acting out complex stories at home, too, as she became aware that her imagination could make all kinds of fun. She was making us coffee, asking if we wanted it hot or cold, with milk or sugar. She was hushing us because her baby doll was sleeping. And she could not stop talking about her friend, the kindergartener, who rode on a real, honest-to-goodness school bus. We also watched, proudly, when she stuck up for herself and brushed off older kids if their play was too intense for her. She seemed to make a huge developmental leap within a couple of weeks of starting childcare.

Delivering Valentines to her friends

I’ve also noticed that the friend that Cee talks about the most at home, and the one that she hugs the longest when she says goodbye, is a little girl just 2 weeks younger than her. I’ve witnessed them engaged in complex, interactive play together, where neither is the obvious leader or follower. And while I think Cee is doing great in her mixed-age childcare setting, I wonder if I’ll feel differently about this as she gets older. We could keep her at Amanda’s through kindergarten, but when she’s four or five, will hanging out with toddlers be a bore? Should we consider moving Cee to a preschool where she’ll be surrounded by kids her own age? Is mixed-age grouping helpful for preschoolers, or does it slow them down?

Child development experts have debated the merits of mixed- vs. single-age classrooms since the 1930s. Soviet psychologist Lev Vygotsky was an early proponent of mixed-age learning. He thought that the best way for children to learn was through interaction with older, more competent children who could set good examples for cognition and behavior. According to Vygotsky, hanging out with older children pushes younger ones to learn more than they would be capable of learning on their own. In contrast, Jean Piaget, well-known for his theory of cognitive development, believed that children learn best from peers close in age, well-matched in knowledge, skills, and power. With their peers, children can learn to solve problems organically and to practice playing a variety of roles in solving them, rather than always being guided in how to fix things by an older, more experienced child.

Decades of research haven’t been able to clarify whether mixed-age or single-age classrooms make for better preschool environments. Some studies point to benefits of mixed-age grouping, similar to my own observations. Toddlers who have a chance to play with slightly older kids seem to develop more complex language and types of play (Rothstein-Fisch and Howes 1988; Howes and Farver 1987). And in mixed-age classes, older children have opportunities to practice leadership skills and prosocial behaviors, like helping and sharing (Derscheid 1997; French et al. 1986).

Mixed-age classrooms might also have the benefit of muddying age-based comparisons and competition between children – a good thing, I think –  because there is likely to be a naturally wide range of skills and abilities within the class. As a parent, it is hard to fret that your child is behind the curve if there is no real curve. And teachers might be more likely to work with children where they are, focusing on their individual interests and skills rather than expecting the entire class to move linearly through milestones and achievements. Plus, mixed-age classes make it logistically easier to keep the same teacher with a group of kids for several years, maybe decreasing the stress of abrupt classroom and teacher changes for young kids (Whaley and Kaplan 1992).

However, there is also a fair amount of research pointing to problems with mixed-ages, the main concern being that mixed-age classrooms don’t challenge older children enough. For example, one study found that older children in a mixed-age setting were more likely to be hanging out solo, detaching themselves from the rest of the group, than if they were in a class of closer peers (Urberg and Kaplan 1986). Another found that older children were less likely to take part in teacher-led activities in mixed-age groups, probably because it is challenging for teachers to plan activities that are appropriate for a broad range in ages (Goldman 1981).

We’re starting to see a pattern emerging here, right? Mixed-age classrooms seem to generally work well for the younger kids in the group, but the older kids might be missing out. A 2002 study headed by Adam Winsler of George Mason University illustrated exactly this pattern in a university-affiliated child development center with morning-only care and a child-centered, play-based philosophy. This study was unique in that it took advantage of the center’s switch from single-age to combined 3- and 4-year-old classrooms, tracking children’s behavioral and cognitive development during the years before and after the switch. Research assistants carefully observed the minute-to-minute activities of the children. Were they engaged in a focused, goal-directed activity? How long was their attention focused on that activity? Were they working alone, with another child (close or distant in age), or with a teacher? A limitation of this design (and most studies like it) is that it can’t be blinded, so it’s possible that the researchers were just seeing what they wanted to see. However, that specific checklist of observations decreases the chances of this being a problem.

How did the children’s behavior differ between single-age and mixed-age classrooms in Winsler’s study? When kids were in single-age classrooms, the 4-year-olds were more “on-task” and had longer attention spans for a given activity. They also spent more time with their peers and less time with their teachers compared to the 3-year-olds. But when those two age groups were combined into one room, the age differences essentially disappeared. The 3-year-olds started acting more like 4-year-olds, and the 4-year-olds started acting more like 3-year-olds. For example, in single-age classes, the 3- and 4-year-olds were working on goal-directed activities 63% and 85% of the time, respectively. When they were thrown into a room together, those numbers shifted to 71% for the 3-year-olds and 70% for the 4-year-olds.

Reading this study got me thinking about the challenges of figuring out what exactly to measure if we’re trying to quantify effectiveness of a preschool program. This isn’t my field at all, so these study methods were new to me. (And, to be fair, the Winsler study is more than a decade old, and the research methods of the field have no doubt changed since then.)

For example, what exactly does it mean for a 3-year-old to be “goal-directed”? Here’s how the Winsler paper described it:

“Goal-directed activity was defined as behavior by the child, which appeared focused, organized, and had an identifiable goal or end point to the activity… Examples of goal-directed activity… included building a structure out of Legos or some other assembly/construction materials, doing a puzzle, playing a game with rules, or engaging in an organized make-believe episode of ‘house.’ Nonexplicitly goal-directed behavior in this context included, for example, aimless wandering around the classroom, looking on into another group’s activity, repeatedly spinning a puzzle piece around one’s finger for the apparent ‘fun of it,’ and making a transition between one activity and another.”

I get that we like to see kids focused on a project or game. Learning is more obvious in this context. But I wouldn’t disregard daydreaming as unproductive activity; science is revealing that daydreaming is actually an indicator of a busy and active brain. And who’s to say that a child isn’t learning a great deal from observing other kids? Or that repeatedly spinning a puzzle piece around one’s finger isn’t the beginning of a passion for physics? Maybe being an older kid in a mixed-age group means spending less time on “school-readiness” activities and more time playing alone and developing empathy and leadership, but how much does each of these contribute to how your kid will ultimately get along in the world? How much do we really know about how kids should and shouldn’t be spending their time?

To their credit, Winsler et al. make this exact point:

“The finding that 3-year-old children in MA [mixed-age] groupings are stretched a bit behaviorally while this is not the case for the 4-year-olds would appear to give some support to the fears reported by parents who typically prefer their child to be the younger child in MA settings (Lloyd, 1999). However, it is important to point out that these data (nor the data from any other studies for that matter) do not imply that MA classrooms are bad for the older children in such classrooms. First of all, it is not clear that the outcome variables in the present study on which the age-by-age-composition interaction effects were observed (sustained activity, goal-directed activity, time spent with peers and a teacher) are predictive of child competence and later positive outcomes. Secondly, there could easily be other advantages to MA grouping for the older children (like the diversity of social experience afforded by increased gender and age desegregation or other social climate variables not explored in the present study) that balance or outweigh these findings.”

In other words, it is difficult for science to capture all the factors that might impact your child’s experience in preschool. Of the factors that we can capture, we’re not even sure how important they really are. Labeling one type of classroom as “better” and another as “worse” is bound to be a gross oversimplification. What the research does show is that the age composition of a classroom impacts how children spend their time, for better or for worse.

This research, with all its limitations, can be useful for informing education policy. For example, more 3-year-olds are now entering Head Start programs. Should Head Start classrooms mix 3- and 4-year-olds together, or should they keep them separate? If they’re mixed, then the research in this field can help highlight important aspects of teacher training, such as strategies for ensuring that children of all ages and abilities are engaged and challenged in their work. A couple of recent studies have looked specifically at the Head Start question, and again, found mixed results (Moller et al. 2008; Bell et al. 2013). There’s a lot more to learn about age composition in preschool classrooms.

Clearly “goal-directed”: Collecting eggs at Amanda’s

I don’t know whether we’ll move Cee to a more traditional preschool before kindergarten. But I like knowing about this research, because it will keep me thinking as I watch her grow up among her dear friends. I’ll be watching to see if she seems bored or disengaged from the group. But there are other factors that will play into our decision, like the stress of moving to a new school and her attachment to Amanda. For Cee, my priorities are that she be in a place where she is happy and feels safe, has lots of opportunities to play solo or with other kids, and that she grows her confidence in social situations. And to be honest, I’m glad she’s in a place where daydreaming isn’t discouraged.