2.1 Participants and settings

We used convenience sampling to recruit 18 parents (or legal guardians) of 7–12-year-olds residing in Southern Alberta, predominantly located near the cities of Lethbridge (population ~100,000) and Calgary (population ~1.3 million). We initially recruited participants through word-of-mouth and posters displayed at the university campus, in local businesses (e.g., coffee shops, YMCA), in a health clinic, and on social media platforms (e.g., Facebook, Twitter). Only one parent per family was interviewed and asked questions regarding one of their children. These methodological choices helped recruit participants who provided a diversity of perspectives. If participants had more than one child in this age range, they were asked to answer questions based on the child whose name came first in alphabetical order. We encouraged participants to refer other eligible parents by sharing the recruitment poster. Informed consent was provided electronically through the DocuSign platform, and the study protocol was approved by the university research ethics board. Participants were provided a $20 gift certificate for the interview and another $20 certificate if they took part in the optional photovoice. Data were collected in the summer and fall of 2022.

2.2 Procedures

Semi-structured interviews, lasting 30–60 minutes, were conducted by a bilingual (English and Spanish) and experienced interviewer. Interviews were conducted online via Zoom, which facilitated the recruitment of participants living in rural areas and avoided travel costs. Interviews began with a preamble introducing the concept of IM. One interview was done in Spanish, and all others were done in English. Interviews were audio-recorded and transcribed verbatim by a research assistant. The Spanish interview was transcribed in English by the interviewer.

The interview guide (Supplementary File 1) included three sections. First, participants were asked general questions about household characteristics that might influence their child’s IM (e.g., parent age, gender, and education; household income; other adults in the household). Second, short response questions were asked to assess the child’s IM based on Hillman, Adams and Whitelegg’ (1990) mobility licenses, which were shown to be valid in a previous Canadian study (Larouche et al, 2017). These assess permission for the child to walk or bike home from school, go to other places within walking distance, cross main roads, go out after dark, cycle on main roads, and travel on local buses on their own. These questions are usually administered in a “yes” or “no” format, and the number of “yes” responses can be tallied to obtain an IM index ranging from zero to six (Larouche et al, 2017). Finally, open-ended questions were asked to explore the factors that encourage or discourage parents from granting their child IM at all levels of influence of the social-ecological model.

In line with social cognitive theory, we asked parents about their confidence in their child’s ability to safely move around their neighborhood without adult supervision. After asking the question in qualitative terms, we asked parents to rate their confidence on a scale ranging from 0 to 10, with 0 representing no confidence and 10 representing full confidence. Then we asked parents about factors that could affect their confidence, including specific questions about mobile phones and dog ownership.

Following the interview, parents were invited to complete an optional photovoice activity. Participants who agreed were provided with practical guidelines for taking photos ethically, based on the recommendations of other researchers and the University of Pennsylvania Institutional Review Board (Bugos et al, 2014). Participants were instructed to take photographs of locations in their neighborhood that either encouraged or discouraged their child’s IM, and to email these images to the senior author, along with a brief description of the location and its significance.

Photovoice is an engaging method that allows participants to express their community’s strengths and concerns through relevant photographs (Wang and Burris, 1997). This technique has been used in previous studies focusing on AT among children (Fusco et al, 2012) and adults (Bhandal and Noonan, 2022; Guell and Ogilvie, 2015). It is a complementary method that can help researchers gather additional data about participant experiences, generating new insights and producing more detailed results (Bhandal and Noonan, 2022; Fusco et al, 2012; Guell and Ogilvie, 2015). The combination of interviews and photovoice exemplifies methodological triangulation and can enhance the rigor of qualitative studies (Bekhet and Zauszniewski, 2012).

2.3 Analysis

The interview transcripts were analyzed with the “framework” approach developed by Ritchie and Spencer (1994), whose ontological perspective is that reality exists independently of individuals’ subjective understanding, but is only accessible through individuals’ perception, discourse, and interpretations (Baldwin and Bick, 2021; Ritchie, Spencer and O’Connor, 2003). It comprises five steps: (1) familiarization with the data, (2) identification of a thematic framework, (3) indexing (i.e., coding) of the interview transcripts, (4) charting, and (5) mapping and interpretation (Ritchie and Spencer, 1994). As described by Ritchie and Spencer (1994), framework analysis is a pragmatic method that allows both themes and sub-themes identified a priori or emerging from the data (i.e., emergent themes). In our case, a priori themes included constructs from the social-ecological model and the social cognitive theory. Four research team members (two women and two men of different ethnicities) were involved in the first two steps, providing a diversity of perspectives. They read the first 10 interviews and identified themes/sub-themes independently before a consensus meeting was held to generate the first draft of the framework. Two team members (the senior author and a research assistant) independently indexed all transcripts in Microsoft Word before meeting to compare their coding and resolve any disagreements by consensus. When needed, they amended the interview framework to capture new emergent themes/sub-themes and/or add nuances to existing ones. Then, the research assistant compiled all coded extracts in a Microsoft Excel document where each sheet provided a compilation of responses to each interview question. Previous research using framework analysis illustrates that Microsoft Excel can be used efficiently for this purpose (Larouche et al, 2012; Swallow, Newton and Van Lottum, 2003). The spreadsheet was reviewed by the senior author, who led the interpretation of the findings and invited the interviewer and research assistants to provide input on interpretations, tables, as well as the selection of quotes to exemplify the different sub-themes.

The senior author also performed descriptive statistics (i.e., frequencies, mean, and standard deviation) on the closed-ended questions to summarize participants’ sociodemographic characteristics, their children’s IM, and parental confidence in their child’s ability to move safely around their neighborhood without adult supervision. The frequency of sub-themes was tabulated to help distinguish common and uncommon themes (Larouche et al, 2012; Lorenc et al, 2008).

3.1 Intrapersonal factors

Table 1 summarizes the factors that were positively and negatively associated with parental confidence in granting IM to their child. The quotes in Table 1 illustrate that parents’ confidence often stemmed from their perception of their child’s characteristics, including maturity, ability to navigate traffic, ask for help, and abide by rules or the law. While many parents attributed their confidence to their child’s abilities, some emphasized the complexity of this confidence. For instance, a mother of a 10-year-old girl pointed out,

“My concern is not with her decision-making, my concern is that sometimes even as an adult, things suddenly happen. My concern is with those unanticipated things; if I am not there obviously, I can’t be, even when I’m there sometimes, weird things happen all of a sudden. I will say we take a risk when we allow independent mobility, but we take a risk when we don’t also, so it is all about weighing those things out.”

This statement points to the complexity of decision-making and challenges the idea that children’s safety is guaranteed in the presence of an adult.

3.2 Interpersonal factors

Parents discussed how they aim to ensure their child’s safety by placing limits and restrictions on their child’s IM (Table 2). Some parents set specific boundaries, including knowing the child’s whereabouts, enforcing curfews when it gets dark, restricting access to certain areas like rivers, and limiting the distance their child can go alone. Others assess IM on a case-by-case basis, such as a mother of a 10-year-old boy who mentioned, “We kind of tailor it [IM] to each child’s personalities and what they feel safe with. Ultimately as a parent, I want my child to be independent, but I also do not want them to be afraid to have that mobility if it is worrisome and starts affecting mental health.” Again, this suggests that parents consider their children’s maturity and personal characteristics when setting boundaries. The quotes in Table 2 suggest parental decisions about their child’s IM are shaped by their relationships, and the child’s ability to conform to parental expectations.

3.3 Community factors

Community factors, such as social norms and perceived views of other parents, can also influence a parent’s willingness to grant their children IM. As highlighted in Table 4, parents had diverging perceptions about whether other parents are supportive or restrictive of their child’s IM. For example, two mothers of 10-year-old boys said, respectively, “I also have neighbours where everybody knows each other, we are meeting on the sidewalks to chat and talk so there is a lot of that high degree of contact between neighbours” versus “I can’t say I know everybody in my neighbourhood, like their opinions but I have to admit I do not see a lot of children running up and down the street.” These statements highlight the concepts of observational learning and how social modeling from the social cognitive theory (Bandura, 1986; Bandura, 1998) may influence parents’ perceptions of what is acceptable or safe for their own children. Three parents perceived a decline in IM compared to when they were kids, and one of them discussed that one contributing factor is that many parents are still at work when the school day finishes, so they register their child in after-school activities where they can pick them up after work. In contrast, a mother of an 11-year-old boy observed increased IM since the pandemic: “Now I do see more kids in our neighbourhood together on their bikes or in the parks. Before the pandemic, I never saw any of that”.

3.4 Built environment

At the built environment level, many parents discussed that access to well-maintained pathways and bike paths and short distances to destinations such as parks, playgrounds, schools, and shops supported IM (Table 5). All photovoice participants included at least one picture of parks or pathways separated from traffic. One mother provided a picture of a pathway that connects to the playground her 8-year-old son usually goes to. Furthermore, two participants included at least one picture showing supportive crosswalk infrastructure for pedestrians. A mother of a 10-year-old boy provided three pictures of flashing lights at an intersection and added, “flashing lights increase my confidence that my son will be able to cross the road safely and that drivers will be able to see him. Not only do the lights get the attention of drivers, but it also makes my son stop and be more aware of his surroundings when he is crossing a major road”. Additionally, low traffic volumes, traffic calming measures, the presence of other responsible adults, and the availability of other children in the neighborhood were identified by parents as elements that can foster IM. A mother of an 8-year-old girl exemplified these factors by stating,

“I think physically it is set up quite well [the built environment] [with] multiple playgrounds, pathways through the ravine in the center of the community, there are two schools in the neighbourhood, there are shops we can walk to, so I like it for those reasons […] There are traffic calming things like speed bumps in front of the schools. It is pretty good physical feature wise.”

A mother of an 8-year-old girl provided a picture of speed bumps on their street and noted that this helps slow traffic, but added, “however, you will note there are no sidewalks, and this is a barrier for young children”. She also provided images of street segments with and without crosswalks and added that “Crossing neighbourhood roads is often more challenging than major ones as there are no crosswalks and my daughter doesn’t have much road sense! She will happily walk out from behind a parked car.” Her contribution highlights interactions between built environment features and her daughter’s behavior that can affect her willingness to grant IM.

Other parents reported that heavy traffic and dangerous drivers who disregard traffic rules discouraged them from granting IM to their child. Perceived high crime rates, particularly those related to drugs and harassment, were emphasized as deterrents as well. For instance, a mother of a 10-year-old girl pointed out, “We have had a lot of problems with needles, persons who are definitely under the influence or overdosing in our neighbourhood, even on our lawn.” Such safety concerns undermined parents’ support for IM. Neighborhood demographics also affect parents’ willingness to grant IM. For example, some parents were concerned that other children and youth could have a bad influence on their child. Other parents worried about a lack of social cohesion and insufficient opportunities for their child to interact with peers. The social-ecological model recognizes the importance of interactions between levels of influence (Sallis et al, 2006) and the views discussed in this paragraph suggest interactions between the social and built environment levels.

Willingness to grant IM varied widely even among parent respondents with children of the same age. Figure 1 displays contrasting opinions about IM from a father of a 10-year-old girl and a mother of a 10-year-old boy. This comparison illustrates how social and built environment factors can affect parental perceptions of IM.

Figure 1

Contrasting opinions about IM. The supportive parent was the mother of a 10-year-old boy, and the restrictive parent was the father of a 10-year-old girl.

3.5 Government policies

Government policies (Table 6), representing the broadest level of influence within the social-ecological model, can significantly impact the built environment factors discussed. This was evident in parents’ comments expressing their desire for governments to implement speed limits and stop signs, and to ensure sufficient police presence to enforce these traffic regulations. Several parents attributed the lack of playgrounds, cycling and walking infrastructure, and enforcement of traffic and safety rules in their area to inadequate policies. One mother of a 10-year-old boy encapsulated these concerns by stating, “I think a big factor [why children’s IM has decreased] is accessibility. We do not live in walkable communities. I do not think we build them… there are not enough bike lanes and paths, there aren’t people obeying the rules of the road…, [and] signages are terrible.” These quotes emphasize how policies prioritizing motorized traffic can create environments that deter parents from granting IM to their children. In contrast, three parents did not believe government policies played a significant role in children’s IM, with one mother of an 8-year-old girl noting that the government does not “have a big influence” and a father of an 11-year-old girl arguing that this is not something that governments pay attention to. The differing opinions may reflect various perspectives about the role and responsibilities of governments at different levels (e.g., municipal, provincial, federal).

4.1 Interpersonal level

Our results suggest that the more confident parents are in their child’s ability to move around unsupervised, the more likely they are to grant IM. These findings are consistent with social cognitive theory (Bandura, 1986; Bandura, 1998) as well as Riazi and colleagues’ (2022) systematic review. Many parents discussed that their confidence is affected by characteristics of their child (e.g., maturity, awareness of traffic) and the broader environment. Our findings align with those of Johansson (2006) who found that Swedish parents’ attitude toward their child’s independent travel depended on characteristics of the child, such as age, maturity, and the perceived need to supervise their child. Johansson (2006) also reported that, although parents gave high ratings to their child’s maturity, they still expressed a strong need to protect their child. Similarly, parents interviewed in our study reported high levels of parental confidence and relatively low levels of IM. This apparent discrepancy may reflect the role of social norms, which was discussed by some interviewees. Dowling (2000) also discussed how, in suburban Sydney (Australia), driving children to school and other places was viewed as “good parenting.”

Moreover, as encapsulated in the quote in our article’s title, some parents spoke about allowing IM as taking a “risk.” A randomized controlled trial has shown that parents’ risk tolerance can be increased through a risk reframing intervention (Brussoni et al, 2021). Risk reframing can help parents recognize the difference between risks (which children can evaluate and decide on a course of action) and hazards (which children cannot assess for themselves and have no clear benefit) while recognizing the developmental benefits of “risky play”—a concept that can encompass IM (Beaulieu and Beno, 2024; Brussoni et al, 2021). This approach could be adapted to promote IM while targeting key correlates of risk tolerance, which, according to a national study, include parental concerns about crime and traffic safety as well as socio-demographic factors (Faulkner et al, 2025).

When asked about the factors that influence their confidence, many parents discussed that their child lacks awareness or skills to navigate traffic. While the provision of pedestrian safety education programs in schools may seem warranted, a systematic review of randomized controlled trials was unable to find evidence that such interventions reduce the risk of road injuries (Duperrex, Bunn and Roberts, 2002). Although they may be well intended, these interventions may unintentionally contribute to blaming the victims in cases of road crashes, such as faulting individuals for not wearing high-visibility clothing (Roberts and Edwards, 2010; te Brömmelstroet, 2024; Walker, Tapp and Davis, 2023). However, such interventions may be popular among parents and policymakers because they do not challenge the hegemony of motor vehicles (Roberts et al, 2010; Walker, Tapp and Davis, 2023). Indeed, over the last century, there has been a cultural shift toward “motonormativity,” which is defined as “a shared bias whereby people judge motorised mobility differently to other comparable topics” (Walker and te Brömmelstroet, 2025, p.1). This widespread bias may explain the disproportionate focus on lecturing vulnerable road users rather than focusing on environmental and policy change that can directly improve road safety (Roberts, 2010; te Brömmelstroet, 2024; Walker, Tapp and Davis, 2023). Examples of the latter include the adoption of Safe Routes to School programs, low traffic neighborhoods, traffic calming, and lower speed limits—all of which have been shown to reduce road injuries (Laverty, Aldred and Goodman, 2021; Roberts, 2010; Tranter, 2018). Broader adoption of such interventions could reduce parental concerns about traffic and potentially increase IM and AT (Tranter, 2018).

Parental support is another interpersonal factor that influences IM. We found that parents used several strategies to support and encourage their child’s IM such as dog ownership and mobile phone use. Previous studies have associated dog ownership with increased IM (Christian et al, 2014: Christian et al, 2016; Riazi et al, 2022), leading to the suggestion that dogs may be IM’s “best friend” (Riazi et al, 2019b). However, our findings provide a more nuanced perspective as other parents noted that effectiveness depends on the dog’s behavior, training, and size. For example, some parents said owning a small or untrained dog did not support their child’s IM. Similarly, we noted that mobile phones can both promote and inhibit IM. Some parents appreciate the security provided by mobile phone usage, enabling easy contact with their child, in line with previous research indicating that mobile phones can increase children’s IM (Larouche et al, 2023; Riazi et al, 2022). However, researchers cautioned that mobile phones could undermine the essence of IM, as parents can track their child’s movements (Larouche et al, 2023; Malone, 2007). Some parents also highlighted issues of children using phones for the “wrong reasons,” such as gaming. To mitigate the drawbacks of mobile phone use, some parents prefer using walkie-talkies, which offer a sense of security without tracking capability, allowing the benefits of IM to be preserved while maintaining parental peace of mind.

4.2 Community level

We found that social norms and perceived views of other parents regarding IM can significantly influence whether parents allow their children to engage in IM. Parents noted a considerable variation in attitudes and social norms within their community, with some discussing living in a safe and supportive environment, whereas others expressed concerns related to drug use (in the context of an opioid crisis in Canada; Belzak and Halverson, 2018) and/or that other parents in their neighborhood are generally unsupportive. The perspective of the father of a 10-year-old who focused on the risk of getting into trouble exemplifies how public spaces are often perceived as a place for adults, and when unsupervised youth occupy these spaces just to spend time together, there is an assumption that they must be “bad” kids that may have a bad influence on others (Iveson, 2006; Valentine, 1996). The routine exclusion of youth from urban planning processes can result in a lack of appropriate places for them, hence greater adoption of programs such as Child Friendly Cities could help support IM (Gilbert et al, 2018; Gleeson and Sipe, 2006).

Interestingly, Mitra and colleagues (2014, p.3408) observed that, when parents perceived that “people are out and about, talking and doing things with one another,” their children had greater IM, but this association was no longer significant after controlling for parents’ attitudes. Parent-perceived social norms could also be influenced by their attitude and upbringings. Multiple recent studies illustrate that the older a parent was when granted IM as a child, the more they restricted their own child’s IM (Gilbert et al, 2023; Hecker et al, 2024; Pelletier, Cornish and Sanders, 2024). Thus, additional efforts should promote IM, engage communities to address barriers to IM, and shift social norms toward greater acceptance and encouragement of children’s IM.

4.3 Built environment

Parents identified several built environment factors that influence their child’s IM. Proximity to parks, playgrounds, schools, and shops along with having well-maintained infrastructure for walking and cycling were often mentioned as supportive factors. Conversely, heavy traffic was identified as a discouraging factor. These factors were so important to one parent that they relocated specifically to provide their children with better infrastructure to increase IM. Interestingly, in their systematic review, Riazi and colleagues (2022) found that destination variables (e.g., walking or cycling infrastructure, greenspace) had no association with IM. However, they identified parental concerns around traffic to be consistently negatively associated with children’s IM. The discrepancies in findings warrant further research to explore associations of both perceived and objectively measured built environment features and IM. Alternatively, based on the social-ecological model, parental perceptions could mediate the effect of built environment attributes on IM and AT (Bauman et al, 2012), suggesting a need for greater use of mediation analyses.

4.4 Government policies

Parents identified several policies that could impact different levels of the social-ecological model, particularly the built environment. For instance, several parents believed that the government should enforce traffic rules more strictly, develop additional playgrounds, as well as cycling and walking infrastructure. Price and Reis (2010) emphasized the importance of these factors by discussing Vancouver’s “Living First” strategy, which focused on creating child-friendly neighborhoods that encourage IM and AT. Additionally, Shaw and colleagues (2015) compared IM levels across 16 countries and found that those with the highest levels often had legislation promoting walking and cycling (Shaw et al, 2015). Collectively, these findings underscore the importance of policies that target different levels of influence on IM. Such policies may not only benefit children but also enhance the quality of life for all residents (Riazi and Faulkner, 2018), especially given that policies can have lasting and widespread impacts (Bauman et al, 2012). Policymakers, urban planners, and transportation engineers can work together to address traffic safety issues as has been done in Safe Routes to School programs in Denmark and the United States, which have been successful in supporting AT while reducing road injuries (Appleyard, 2003; DiMaggio, Frangos and Li, 2016; McDonald et al, 2014). While Safe Routes to School programs have overlooked IM as an indicator of impact, one may hypothesize that the systematic adoption of this approach contributes to the higher levels of IM observed among Danish children compared to most other countries (Shaw et al, 2015).

4.5 Strengths and limitations

One limitation of the study is that, because interviews were conducted at a single time point, we were unable to assess changes over time. The use of a small convenience sample of parents in Southern Alberta limits the generalizability of the results. While qualitative research does not emphasize generalizability, it can provide rich contextual information for stakeholders (e.g., policymakers, practitioners). It is worth noting that 88% of our participants reported having university education and this proportion is much higher compared to the 2021 census, in which 31.1% of 25- to 64-year-olds reported holding at least a bachelor’s degree (Statistics Canada, 2022). Given that the systematic review by Riazi et al (2022) identified no association between parental education and IM, it is unclear how the overrepresentation of highly educated parents affected our findings. Additionally, the applicability of some IM licenses was constrained by contextual factors, such as the lack of transit options in rural areas. We cannot tell how many participants lived in urban and rural areas, as this information was not collected. The underrepresentation of fathers in the sample is typical of parenting studies (Davison et al, 2016); however, the results may be biased toward maternal perspectives on children’s IM. Finally, subjective methods such as interviews and photovoice are vulnerable to social desirability bias.

Despite these limitations, the study has several strengths. The combination of interviews and photovoice provided complementary insights into the factors associated with children’s IM. Additionally, triangulation of methods, researchers, and theories enhances analytic rigor. Moreover, the study’s pragmatic focus on the strategies used by parents to manage their children’s IM offers practical insights that can inform the development of future interventions.