Risky play helps children develop real-world safety skills, new virtual reality research suggests

A recent study published in the Journal of Environmental Psychology suggests that children who engage in risky play tend to develop better risk management skills for real-world situations. The research provides evidence that taking physical chances during playtime helps children learn to navigate higher-consequence hazards, such as crossing a busy street, with greater efficiency. Children who took more risks on the playground made safe decisions more quickly when navigating a simulated busy intersection.

Risky play refers to activities that involve a sense of uncertainty and physical thrill. Examples include climbing high structures, moving at fast speeds, or exploring a neighborhood without adult supervision. Adults often view these activities as hazardous. But allowing children to test their physical limits tends to foster resilience, physical coordination, and emotional growth.

When children encounter physical risks, they engage in a framework of behaviors known as the dynamic risk management model. The first step in this model is an emotional process called risk willingness, which involves a child’s natural interest in taking on a physical challenge. They also practice risk assessment, a cognitive process of evaluating the danger before acting. Finally, they engage in risk handling, which refers to the actual physical movements taken to navigate the situation.

Over time, these repeated cycles of perceiving a challenge and taking physical action help children build a mental map of their capabilities. This continuous learning cycle is known as the perception-action loop. Through repeated trial and error, children learn to interpret environmental clues and adjust their physical movements accordingly. Environmental psychologists often refer to these environmental clues as affordances, which are the physical opportunities for action that a specific space provides relative to a child’s unique abilities.

Past research supports the idea that risky play builds general competence in traditional play settings. However, it remained unknown whether these skills transfer to completely different, non-play scenarios. The current research was designed to test whether the benefits of childhood risk-taking during play actually translate into practical safety skills, like pedestrian navigation.

Free daily newsletter

Parents, schools, and policymakers have spent decades trying to make childhood safer by removing risk. But researchers theorize that if children never get to practice assessing and confronting small, manageable dangers, they may not develop the judgment they need for bigger ones.

Mariana Brussoni, a professor at the University of British Columbia department of pediatrics and school of population and public health, director of the Human Early Learning Partnership, and researcher at BC Children’s Hospital Research Institute, helped lead the research. “This risky play is a fundamental way that children learn about the world, about themselves and how to keep themselves safe in diverse situations,” Brussoni said.

The authors also wanted to explore how different cultural environments might shape a child’s willingness to take physical chances. For instance, Norwegian culture and national education policies actively encourage outdoor independence and risk-tolerant play. By contrast, Canadian environments tend to feature more restricted, supervised childhoods. To investigate this, an international team of scientists from the University of British Columbia, Queen Maud University College of Early Childhood Education in Norway, and Colorado State University collaborated on the project.

To safely observe children navigating hazards, the researchers utilized immersive virtual reality technology developed by the ViRMa project at Queen Maud University College. This allowed the scientists to place children in simulated risky environments without exposing them to actual bodily harm. A notable 85 percent of the children reported that the virtual reality environments felt realistic to them.

This technological approach was necessary because of the obvious dangers of testing pedestrian behavior in the real world. “I don’t think there’s any ethics board in the world that would allow you to throw children into traffic to see how they did,” Brussoni said. Brussoni authored the study alongside Ellen Beate Hansen Sandseter, a researcher at Queen Maud University College.

“We really haven’t been able to properly test this hypothesis before we had access to these kinds of technologies,” Brussoni added. The sample consisted of 424 children between the ages of seven and eleven. Specifically, the group included 361 children from Norway and 63 children from Canada.

Each child completed two separate virtual reality tasks while wearing a specialized headset and moving around a real gymnasium. The first task was a playground scenario designed to measure risk willingness and the likelihood of failure. Children explored a virtual structure made of balance beams and small freestanding pillars. The simulated heights varied across the structure, reaching up to 1.5 meters at the highest point.

The researchers measured several variables to calculate an overall risk willingness score. They tracked how fast the children moved, how much time they spent in the most dangerous elevated sections, and how often they ventured onto the tricky freestanding pillars. The children were given a three-minute window to explore the virtual environment as they saw fit. If a child lost their simulated balance and fell, it was recorded as a play failure.

The second task simulated an urban pedestrian traffic environment to measure risk management outside of a play context. Children had to decide when it was safe to cross a virtual bicycle path and a simulated street with oncoming traffic. The virtual vehicles moved at a constant speed of five meters per second. The traffic density started high and gradually decreased over time, making the crossing easier the longer the child waited.

The scientists recorded the children’s assessment time across the six pedestrian tasks. This measurement captured how long the children waited before deciding to cross the road. On average, the participants spent about 107 seconds evaluating the traffic across all trials. The researchers also tracked collisions and near-misses, defined as crossing within 0.8 seconds of a moving vehicle. Across the sample, 27 percent of the children managed to cross safely in all six trials.

The researchers found that Norwegian children demonstrated significantly higher risk willingness during the playground task than their Canadian peers. This pattern remained true even after adjusting for differences in age and sex. The Canadian participants were slightly older on average, but they still displayed more risk-averse behavior. Older children and boys generally exhibited higher risk willingness across both cultural groups.

Among all participants, having a higher risk willingness was associated with a greater likelihood of falling off the virtual playground equipment. Overall, 21 percent of the children experienced a simulated fall during the playground scenario. Statistical models showed that a unit increase in a child’s risk willingness score raised the odds of a simulated fall by 78 percent. The researchers suggest that falling, stumbling, and trying again teaches children what they are capable of, where their limits are, and how to adjust.

Interestingly, this increased willingness to take risks during play translated to more efficient decision-making in the higher-stakes traffic scenario. Children who showed higher risk willingness in the playground spent significantly less time evaluating the traffic before deciding to cross the street. For example, a child with the highest risk willingness score spent about 68 seconds less assessing the traffic than a child with the lowest score.

Most importantly, this faster decision-making did not result in more pedestrian accidents. The researchers observed that higher risk willingness was not associated with an increase in dangerous choices, such as collisions or near-misses with the virtual vehicles. The children with higher risk willingness simply processed the environmental information faster, read the situation, and made safe crossing decisions more efficiently.

While the study provides evidence for the benefits of risky play, readers should avoid interpreting these findings as a recommendation to expose children to extreme or unmanageable dangers. The concept of risky play revolves around manageable challenges that match a child’s growing physical capabilities, rather than reckless endangerment. The findings suggest that the playgrounds we design, and the freedoms we grant or withhold, might be shaping a child’s ability to navigate a complex world long after they leave the swings behind.

“Keeping children safe means letting them take risks,” Brussoni said. To support children’s outdoor risky play, Brussoni points to three main ingredients: time, space, and freedom. She suggests parents carve out real, unstructured time every day and find interesting places to play with other children, rather than relying solely on standard playground equipment that kids quickly outgrow.

Parents should then step back enough to let children actually play, which includes allowing small physical risks that might feel scary to watch. For parents who struggle to resist intervening, Brussoni offers a simple trick. She recommends counting to 17 before saying “be careful.” This brief pause is just long enough to shift from a fear response to a more considered reaction.

For communities, the findings point toward a need to cultivate collective understanding around the importance of risky play and independent mobility for children. This means ensuring there are natural and creative spaces for play that are as safe as necessary, rather than as safe as possible.

The authors note a few limitations to their research. Although a vast majority of the children rated the virtual reality environments as highly realistic, computer simulations cannot perfectly capture the complex physical nuances of the real world. The pedestrian task also used only one trial per difficulty level. This design choice reduced testing fatigue but limited the researchers’ ability to analyze how an individual child’s behavior might vary across multiple identical attempts.

In addition, the virtual reality setup was not designed to accommodate children who use physical mobility aids, such as wheelchairs. This aspect of the methodology limits the ability to generalize the findings to children with different physical disabilities. Scientists will need to adapt these virtual environments to ensure future research includes participants of all physical abilities.

Another notable detail is that the Canadian children were recruited from a specific school known for encouraging outdoor risky play. Because this school is not entirely representative of standard Canadian educational environments, the sample might reflect an unusually high risk tolerance for the region. The true cultural gap in risk willingness between typical Norwegian and Canadian children might be even larger than what the scientists observed in this specific study.

Future research could explore a wider variety of cultural groups and physical scenarios to see if these patterns hold across different populations. Scientists might also develop new technological methods to track children’s behavior in actual real-world settings to verify the results gathered from the virtual reality simulations.

The study, “The developmental importance of risky play: A cross-national virtual reality study,” was authored by Mariana Brussoni, Ellen Beate Hansen Sandseter, Ole Johan Sando, Rasmus Kleppe, Megan Zeni, and Anita Bundy.