One of the primary objectives of this study was to draw comparisons between unescorted (walking) and escorted (walking or driven) children in a sample living within two kilometers from school with regard to compositional and contextual dimensions, and parental AST related perceptual differences. This comparative cross-sectional analysis was designed to produce novel insights into how we may shift school travel mode choice from driving to walking so that children can experience the PA and independent mobility benefits of unescorted AST trips. A second objective was to examine methods and strategies that could increase the likelihood of AST among the escorted drivers.
In terms of demographics, unescorted walkers were significantly older and the families spoke predominantly English at home, and were more likely to live within one kilometer from school. Although previous research has shown mixed findings in terms of the association between age and AST, recent studies have shown age to be significantly associated with children’s independent mobility to/from school , and that older children are less likely to be escorted to/from school by their parents [34, 35]. Our finding appears to imply that there may be an age threshold where parents believe that their child has the cognitive capacity to navigate his/her way to school safely. Results showed that parents of children who are escorted to school (walked or driven) would only allow their child to walk unescorted at the mean ages 11.5 (currently walking escorted), and 11.84 (currently driven), respectively, when the child is approaching junior high school. Any age lower than the threshold may be viewed by parents as being an unsafe age for children to actively travel to school without adult supervision. This is supported in previous literature, which suggests that younger children are at higher risk when exposed to traffic situations  due to their attentional skills and their age-moderated appetite for risk taking [41, 42]. Additionally, parents may feel that an older child possesses a greater sense of agency, and is therefore more capable of advocating for her/his own safety when in the presence of strangers.
The GTHA is composed of the metropolitan areas of the cities of Toronto and Hamilton, Ontario along with the regional municipalities of Durham, Halton, Peel and York. This region has varying built environments spread across traditional downtown, inner and outer suburban areas. Notably, region, at the broad scale applied here, did not differentiate those who walked unescorted from those who were escorted. Rather, a child living within one kilometer from school was more likely to walk unescorted regardless of regional location. We expect, however, that marked difference could emerge with a shift in scale that would permit neighborhood by neighborhood comparison within an individual city or regional municipality.
Distance is consistently found to be negatively associated with AST . Our findings highlight that, even in a sample living within a ‘walkable’ distance from school, distance remains a strong predictor of both travel mode and independent travel. Similarly, Fyhri and Hjorthol  reported that increasing the distance to school from one to two kilometers was associated with a significant decrease in independent mobility. A reduced distance from school may help parents feel more at ease with their child walking unescorted due to a reduction in the time exposed to risk (e.g., fewer streets to cross, fewer opportunities to run into strangers on a short trip). To promote AST, these findings reinforce the fundamental importance of ensuring school catchment areas are proximate to schools [43, 44].
This study also portrayed that the families of children who were unescorted spoke predominantly English at home when compared with families of the children who were escorted to school. Given that there were no differences in terms of household income, this finding may suggest ethnic or cultural differences in independent mobility, in line with previous research showing that minority ethnic children were more restricted in their independence . Future research might be informative in considering cultural variations in perceptions of independent mobility and school travel. Giuliano  noted that understandings about travel behavior have historically focused on Caucasians, with less attention given to the question of how race/ethnicity manifests itself within location decisions and travel behavior. More work is needed to understand the travel motivations, constraints, and travel outcomes of various ethnic groups. Certainly, in the presence of an increasingly complex landscape of diversity within cities and regions, some careful attention (in research addressing travel behavior) should be given to how we define difference, and how difference produces or is produced by travel behavior decisions and outcomes. Data limitations prevent such detailed analysis in this work, where instead, the primary language spoken at home is used as a measure of language background. The authors acknowledge that while this variable relates to culture, it is an incomplete indicator of, for example, race and/or ethnicity [47, 48].
As a novel contribution, we also found that children were more likely to walk unescorted to/from school if their parents agreed to a greater extent that they chose to reside in the current neighborhood in order for their child to practice AST. Those families inclined to walk for transportation in general, or who are interested in their children walking to school may seek out and eventually reside in neighborhoods that are more conducive (e.g., closer) to practicing AST and achieving independent mobility. This suggests that neighborhood self-selection may be an important correlate to consider when studying AST. That is, higher rates of AST for example might not be caused by built environment characteristics that support AST in specific neighbourhoods . Rather, parents who attach importance to a physically active lifestyle might select homes and neighbourhoods that are generally walkable in nature (note in our sample, drivers placed less importance on their child exercising en route to/from school). Based on recent reviews in the AST literature [49, 50], residential self-selection has not been explicitly discussed and future research examining factors influencing AST should control for residential self-selection in their analyses where possible.
In terms of parents’ perceptions and attitudes related to AST, safety concerns clearly differentiated the unescorted and escorted groups. In general, escorting parents had greater concerns about strangers/bullies approaching their child, and parents who drove their children to school were additionally concerned with traffic volume around school. Fear of child abduction or ‘stranger danger’ , neighborhood violence [51, 52], and traffic volume  are all significant barriers to AST and independent mobility in children. Ironically, driving parents also had greater concerns that there were too many cars in the morning around their child’s school. This concern likely results from the fact that even children that are driven end up pedestrians at the school-end, around the site – and, therefore, experience some exposure to risk that is limited to the destination. Similarly, a recent study found that parents who use their car frequently have higher risk perceptions regarding AST . Parental concerns about traffic safety typically are related to perceptions about the number (traffic volume) and speed of vehicles around the school. In an effort to protect their child, many parents drive their children to and from school. Paradoxically, while they may perceive they are reducing the risk of injury for their own child, they are also contributing to the problem and fear of too many vehicles in the school neighborhood .
Given the importance of age, distance and safety issues as significant correlates of independent mobility, non-infrastructure programs and policies should be developed that provide adult supervision. An often-discussed intervention is the Walking School Bus (WSB). Proposed by David Engwicht in the early 1990s, a WSB entails children walking escorted by an adult to and from school with other children. Similar to a school bus, these adult supervisors act as ‘drivers’ and collect children along the route to school and actively escort them in a timely, safe manner. Although a WSB provides adult escorts, shifting school travel modes from driving to walking with an adult escort undoubtedly enhances greater independent mobility compared to being escorted to school by car. This was Engwicht’s original intention; the purpose of initiating a WSB was to foster greater independent mobility by providing initial parental guidance. He envisioned that children would graduate to full independence once parents gained confidence in their child’s ability to navigate his/her way to school. More recently, Engwicht  reiterated that it is critically important to view WSB schemes as an intermediary step towards a child’s road to full independence. Further, WSB schemes may help establish walking as a normal, daily behavior which promotes walking as a transportation mode of choice for extracurricular activities during a child’s school years and beyond.
Findings from this study support the need to organize WSBs to promote AST, especially for younger students. Parents of driven children were significantly more likely to utilize an organized WSB. A similar measure from this study provided some cause for optimism in that 67% of parent drivers reported that access to a WSB would increase the chances of their child practicing AST while in general parents of children driven to school were also more likely to endorse the appeal of walking to school with their child. However, given the cross-sectional design of this study, causal relationships cannot be assumed between examined variables. For example, it cannot be assumed that providing access to WSB schemes would assist a shift from being chauffeured to school to supervised walking. While the findings indicate that schemes like WSB would appear to be a logical solution in addressing the safety concerns of parents who currently drive their children to school, they are not necessarily common interventions in Canada . There are many challenges in the sustainability of such schemes, such as parent and volunteer compliance, school support, and investment from key stakeholders [58, 59]. Recent research also suggests that WSBs may not be cost-effective . However, this study was limited as cost-effectiveness was examined relative to obesity prevention rather than the broader benefits that more active forms of travel may incur. More research is required in understanding the optimal formation and sustainability of these schemes.
Other strategies might be considered that have the goal of alleviating parental concerns regarding the independent mobility of children. For example, an ‘arrival confirmation’ call/text can be made by school administration after attendance is taken to parents of those who actively transported to school. With increasing number of children owning cellular phones, parents should perhaps program the child’s cell device to enable instant communication with a caregiver with a touch of a key (e.g., ‘1’= mom, ‘2’=dad, ‘3’=police), in case the child is in danger or needs help. Alternatively, the role of a crossing-guard can expand to serve two purposes; 1) to help children cross busy streets in a safe manner or 2) to act as ‘eyes on the street’ as they stroll the streets between the school and surrounding neighborhoods. The identification and promotion of school specific ‘safe’ routes to school that minimize traffic exposure and follow routes with high pedestrian use is another example.
In Canada, strategies like these to increase AST and independent mobility are likely to be embedded within broader School Travel Planning (STP) initiatives (see ). STP is a multi-disciplinary, multi-sectoral intervention that engages key stakeholders (i.e., public health, police officials, municipal planners & traffic engineers, school boards, parents and school administrators) to document and evaluate school travel issues and develop and implement a travel or ‘action’ plan that creates environments that are more conducive for children to practice AST to/from school . Stemming from the action plan, for example, schools could develop a WSB. Other strategies could be adopted to reflect some of the strategies endorsed in the current study by parents. For example, 63% of parents of children driven to school indicated that having crossing guards and marked crossings by the school would increase the likelihood of their child walking to/from school. Parents also reported that having police presence around the school, having school zone cautionary signs posted, having well maintained sidewalks, and seeing slower speeds around the school area would increase the likeliness of their child practicing AST. It is not clear whether such features were absent or already exist at the schools of the surveyed parents.
School travel plans could also call for educational activities to highlight the contribution AST makes to overall daily energy expenditure levels of children, and to promote walking and cycling safety. Our findings suggest that escorting parents are not talking to their children about safety issues nor are they endorsing the trip to school as an opportunity for physical activity. Addressing these issues through parental educational programs may be required, particularly since past educational interventions have helped increase rates of AST [62, 63]. Although AST typically is associated with walking and biking to school, this study explored group differences in terms of walking since only 1% of the sample cycled to school. Future research would be interesting in examining households where children cycle independently to school.