Children's School Commutes in Distinct Urban Neighbourhoods

A Case of Visakhapatnam, India

Abstract

Commuting to school is one of the major activities of children in their daily lives. Over the last two decades, children's school commutes in Indian urban areas have significantly transitioned from active to passive modes. Expansion of school choice policies, new schools and the growth of automobiles are the leading factors of the increased home-to-school distance and diverse commute modes in India. This study investigates the children's school commute patterns in distinct urban neighbourhoods. This study employed a questionnaire-based cross-sectional study in four urban neighbourhoods in Visakhapatnam, India. The neighbourhoods were selected based on location (inner metropolitan, suburban established, suburban isolated and transient neighbourhoods), land use proportion, temporal changes and access to transportation facilities. Chi-square tests were employed to assess the association among the variables, and a multilevel multinomial logistic regression was performed to predict the odds of various commute modes across neighbourhoods. The results show a significant association of neighbourhood-specific factors in children's school commute mode. The threshold of walking and cycling varied among the neighbourhoods. Urban planners and policymakers should focus on pedestrian and cycle-friendly infrastructure in neighbourhood environments to encourage and promote active school commutes.

Introduction

Over the last two decades, India has witnessed a growth of 55.4% in new schools (PIB, 2022). Such expansion was the significant result of policies aiming to provide access to education in India. During this period, widespread school choice opportunities and affordable transport facilities have gradually influenced parental school selection criteria and eased the limits to home-to-school distance. This transition has resulted in more extended school commutes and diverse commute modes in urban and rural areas. On the other hand, India has also witnessed a 727 % growth in registered motor vehicles between 2000 and 2020 (MORTH, 2023). The tremendous increase in automobiles, rapid urbanization, and inadequate pedestrian infrastructure has resulted in hazardous conditions for walking and cycling, which are the driving factors of varied school commutes in many urban areas.

Concurrently, a growing concern in children's school commutes arises from their vulnerability as urban road users. The World Health Organization (WHO) recognizes road traffic accidents as the primary cause of injuries and fatalities in children, and approximately 95 % of such incidents are reported in developing low and middle-income countries (WHO, 2017) . In India, road traffic accidents involving children are reported annually, with 9.4 % of fatalities (Singh, D., Singh et al., 2016) and 17 % of injuries (Tetali, Edwards, Murthy et al., 2016) during school commute. These numbers are much higher than those of the neighbouring country, China, which reports 1.7 % fatalities. In addition, the long school commutes further prevailed by passive commutes (Manjunatha, Uthkarsh et al., 2020; Sidharthan, Bhat et al., 2011), contributing to traffic congestion at urban road intersections during school start and end times (Srinivasan, 2010). The findings of (Sun, W., Guo et al., 2021) study show that the likelihood of traffic congestion on roads rises by 4.5 percentage points by school trips. These impacts further vary based on the characteristics of the school, route, and neighbourhood. The dominance of passive school commutes further contributes to the excess consumption of fossil fuels and the release of toxic environmental pollutants into the environment (Marshall, Wilson et al., 2010). Recent studies also highlight a significant drop in active commutes even for less than a kilometre distance to the school in urban areas (Easton and Ferrari, 2015; McDonald, 2008; Mitra and Buliung, 2015).

Recently, India surpassed China in population and became the most populous country in the World, with nearly 1.42 billion (Hertog, 2023), and school-going children account for more than 265 million. Similar to adults, children are daily road users; however, their travel behaviour entirely differs from adults. It emphasizes a significant concern about understanding children's school commute behaviour in urban areas.

This study investigates the school children's commuting behaviour in diverse neighbourhood settings, considering demographic, socioeconomic, and built-environment characteristics. This study provides insights into the following questions in an Indian urban context.

1) What are the various school commute modes of the children in urban neighbourhoods with distinct spatial and temporal urbanization contexts?

2) Does children's commuting patterns correlate with their demographic, socioeconomic backgrounds and neighbourhood characteristics?

This study assesses the association between self-reported perceptions and objectively measured environmental characteristics of home-school commutes to examine the above questions. The findings apprise contemporary urban neighbourhood concepts related to children's commuting to school

Literature Review

Understanding School Commutes

Commuting to school is one of the major activities of children in their daily lives, and it further refers to the children's commuting behaviour between home and school. Walking, cycling, escorting by another person, taking a school bus, public transportation, or any other vehicle are various modes of transportation. Authors in many studies refer to commuting through walking and cycling as an active mode of commute and by means of motorized vehicles as a passive mode. Globally, several empirical studies state that an active mode of school commute contributes to a child's daily physical activity needs and provides various health benefits, including improved cardiovascular health, increased physical fitness, and a positive impact on mental well-being (Adlakha, Hipp et al., 2018; Lee, 2020; Omura, Hyde et al., 2019) . In addition, it promotes a sense of independence and responsibility among students (Mandic, Williams et al., 2016). In contrast, passive commutes offer leverage to time and distance constraints for school commutes. However, several studies report these commutes are also associated with several health risks like obesity (Sun, Y., Liu et al., 2015), cardiovascular diseases (Seaton, Godden et al., 1995; Timperio, Ball et al., 2006) and other lifestyle disorders (Faulkner, G. E. J., Buliung et al., 2009) to children in the long run.

Factors Affecting School Commutes

Globally, many studies have identified various significant factors influencing travel mode choice decisions for school trips. In the early phases, studies were centred on children's age, sex and family's socioeconomic status as major sociodemographic attributes in exploring the children's school travel behaviour. Subsequent studies were incorporated that, apart from the sociodemographic factors, various environmental factors like Distance (Davison, Werder et al., 2008), urban form (McMillan, 2007), land use, and street design pattern (Bi and Zhang, 2016; Sharmin and Kamruzzaman, 2017), neighbourhood environment around home and school, and route characteristics between home-to-school(Panter, Jones et al., 2010) social and cultural settings (Waygood and Kitamura, 2009) may impact a child's commuting choices. Such studies contend distance is the critical indicator in mode choice; walking and cycling to school tend to decrease, and dependency on vehicular commute may increase while home-to-school distance increases (McDonald, 2008; Mehdizadeh, Mamdoohi et al., 2017; Mitra, 2013). However, the home-to-school distance is not a direct indicator of mode choice; factors like household income, parental education, ownership of vehicles, neighbourhood safety, number of children, and age and sex of the child may have a significant association in commute mode choice.

Child’s Characteristics

Child's age/education level could be one of the driving factors in mode choice; as children grow up, their ability to commute to school becomes more independent. Thus, age is crucial in explaining school commuting choices (Carver, Timperio et al., 2013; Reddy and Pasupuleti, 2024). Many studies say a parent or caregiver mostly accompanies younger children, and a strong association between increased age and walking and cycling to school has been found. Surprisingly, few studies have found no association (Wilson, Coen et al., 2019) or negative association (McMillan, 2005) to the child's commute mode with age. Regarding sex, studies show boys use more active modes than girls, mainly commuting by cycling. This finding has further exercised by age and household factors. (Faulkner, G. E., Richichi et al., 2010) state that up to a certain age, the transportation mode for getting to school is typically determined by parents or guardians within the family. Boys get more freedom to explore the environment outside the home at an earlier age than girls, which also influences commuting behaviour. Up to a certain age, parents or caregivers escort both girls and boys and mode choice is positively influenced by the household characteristics (Marzi and Reimers, 2018).

Household Characteristics

Household characteristics like the number of children, family income, parent's education, occupation, and vehicle ownership reflect on children's school commute mode choices (Mehdizadeh, Mamdoohi et al., 2017). Children from high-income families have found commuting less independently and relying on assisted travel, whereas low-income children prefer active commute modes to school trips. On the other hand, high-income parents tend to choose the best schools rather than the nearest to their house, which prompts their children to drive to the school (Mandic, Sandretto et al., 2023). Numerous factors may account for this depending on the context; parents' occupation is one of the associations with the child's school commute, as parents travel to their work tend to drop and pick up their child from school and primarily working parents choose the schools located in en route to their work location. Studies focused on household composition and school commutes found that children from single-parent families may be accompanied less often due to time constraints (He, 2013; Hsu and Saphores, 2014; Yang, Wang et al., 2018). Additionally, a consistent finding has been noticed in many studies that car ownership within households is a potent travel resource that can exert its effect independently of its connection to social status (Carver, Timperio et al., 2013).

Built-environment Characteristics

A phenomenal work of (McMillan, 2005) on urban form and children's commuting behaviour states that urban form may not be directly associated with mode choice; it may vary with mediating factors like neighbourhood safety, road traffic, and household transportation options and moderating factors like socio/cultural norms, parental attitudes and sociodemographic factors in contribution to parental decision-making on children's school commuting behaviour.

While most such studies examined either self-reported perceptions or objectively measured results on environmental attributes, (Panter, Jones et al., 2010) used a combination of both measures. They identified that sidewalks, pavements and mixed land use around the school environment could be positively associated with active commuting. Thoroughfare traffic roads and the number of intersections along the home-to-school route may negatively affect active commuting.

Perceived Risks on Road Safety

Regarding safety, multiple dimensions were explored in the earlier studies, ranging from the child's specific factors to the neighbourhood environment. Such studies highlight many parents' lack of confidence in their child's ability to travel independently to long commuting distances. In addition, the danger of road traffic, lack of social support, encounters with strangers, and fear of crime were identified as barriers to active school commuting. Road traffic accidents are the primary cause of injuries among children in urban areas in both developing and developed countries. A recent study in India reports that one-sixth of the students experienced traffic-related road injuries while walking or cycling to school (Tetali, Edwards, Murthy et al., 2016).

Methodology

This section first introduces the study setting and then explains the selection of study areas and data collection process, further the variables used in the study. Finally, it concludes with the statistical analysis used in the study.

Study Context

Visakhapatnam is a metropolitan city and major hub for educational, industrial and commercial activities in the state of Andhra Pradesh in India. As per the 2011 census, the city has a population of >1.7 million and is India's 17th most populous city. The metropolitan area is divided into 72 municipal wards, administered by six zones, and spread over 539.95 square kilometres (See Figure 1 A & B). Further, the city has 970 schools and more than 380,000 enrolments. In this study, we used a cross-sectional study and recruited a diverse sample of participants from four distinct neighbourhoods in the metropolitan area. Neighbourhood areas were selected for this study based on specific urban characteristics. Random sampling was used to distribute the instrument in data collection in the selected neighbourhoods. Parent's contact details (household data) were gathered from the ward secretariat (a ward-level governing body under the municipal corporation) of the localities chosen with the assistance of the local representatives, and the questionnaire form was distributed to the households of the respective areas.

Figure 1. A) Map showing the Visakhapatnam urban area in Andhra Pradesh, India. B) Map showing the boundary of city, zones & wards and locations of selected four study areas.

Data and Methods

In this study, we adopted (Wong, C., Zheng et al., 2020) classification for identifying the neighbourhoods to investigate the children's school commuting patterns in urban areas by selecting distinct urban neighbourhoods based on location, distance from the core city area, density, accessibility to transportation facilities, temporal changes (development period). And (Panter, Jones et al., 2010) objectively measured neighbourhood environment variables for school commute assessment. A structured closed-ended questionnaire is administered to collect data on children's commute behaviour in four neighbourhoods. The participants are the parents or caregivers of school children aged 5-15 years. Characteristics of demographic, socioeconomic, neighbourhood environment, school location and commute mode were collected in Visakhapatnam between July 2021- December 2021. A total of 792 observations were received from the selected study areas. The purpose of the research and its objectives were clearly stated in the questionnaire form, and informed consent was obtained from the participants before filling the questionnaire. This study is administered from the neighbourhood perspective; the school's name and distance of the school from home are self-reported by the participants. The study did not mark exact home addresses/locations; however, the on-foot survey by the author documented the variables of neighbourhood built environments and school routes. Figure 1B shows the locations of the four surveyed urban areas in Visakhapatnam.

Observed Built-environment Characteristics of The Selected Neighbourhoods

The observed characteristics of the built environment and the data sources of the four areas are presented in Table 1. The neighbourhoods are selected based on spatial urbanisation and locational factors. The inner metropolitan (MVP colony) is the first developed neighbourhood, followed by suburban established (Steelplant township - a satellite town), suburban isolated (Muralinagar and transient (Arilova colony) neighbourhoods. The inner metropolitan and suburban established areas are completely designed neighbourhoods that share 48.39 % and 49.23% of residential area and are well connected to transportation and urban services. On the other hand, suburban isolated and transient neighbourhoods are mixed with designed and informal settlements, sharing 64.3% and 68.15% of residential areas, respectively. Figure 2 and Figure 3 show the proportion of existing land cover mix of all neighbourhood areas proportion and spatial representation, respectively. An overview of the neighbourhood variables, their description and source of data is mentioned in Table 1.

Figure 2. Existing land cover proportion of four selected neighbourhoods.

Figure 3. Existing Land Cover of selected study areas.

Table 1. Selected neighbourhoods and their spatial characteristics

Variables	Definition	Data Source
Development period	The period of the neighbourhood development.	Local municipal development authority. (GVMC)
Distance of the selected areas from the city centre	Euclidean distance between the City's Major bus terminus to the selected neighbourhoods.	Measured by Google Earth
Access to public transport (bus stops)	The existing number of bus stops in the neighbourhood within a 1-kilometer radius.	Author's field study.
Land ownership	There are two types of land ownership: (1) private individuals and (2) government or society.	Author's field study.
Population density	The number of people (residents) per acre area	Census 2011, India.
Land use	Types of land use parcels within the Neighbourhood.	2021 master plan, GVMC.
Commute route	Streets, Roads, Pavements, and other built-environment parameters	Author's field study.
Selected neighbourhood Type	Characteristics
Inner metropolitan Neighbourhood (MVP Colony) (study area-1)	Close to the urban core (4km from the city centre)    Properly designed township (developed during the 1970-80)    Grid-iron pattern roads and settlements    Well-connected roads, convenient access to public transport and other facilities    Relatively very high land price level    Land owned by individuals and private    Sufficient neighbourhood parks
Suburban Established Neighbourhood (Steel plant township (A Satellite town)) (study area-2)	Located in a suburban area (26km from the city centre)    Properly planned and designed township (developed during the 1980-85)    Radial pattern roads and settlements    Limited accessibility to public transportation    Land owned by society (government)    Sufficient open green spaces and parks
Suburban Isolated Neighbourhood (Murali Nagar) (study area-3)	Located in an urban fringe area (7km from the city centre)    Combination of designed layouts and informal layouts (developed during 2000 and later)    Moderately high land price level    Limited accessibility to public transport but somewhat close to facilities    Land owned by individuals and private    Limited parks in designed areas and no open and green spaces in informal layouts
Transient Neighbourhood (Arilova Colony) (study area-4)	Located in an urban periphery area (9 km from the city centre)    Designed layout for the urban working-class population    Relatively Small land plots and dense residential land.    Grid-iron pattern roads and settlements.    Limited access to public transport and facilities    Land owned by individuals and private    Compact development with no green spaces and open areas.

Statistical Analysis

A sequence of methodological procedures was conducted for analyzing the data. All the statistical tests are conducted by using IBM® SPSS® Statistics Version 27. Overall observations were summarized as frequencies and percentages for each variable. The proportion of children studying within the neighbourhood (school located < 1km) and outside the neighbourhood (school located > 1km) are calculated for each neighbourhood (Figure 4). Our study's outcome variable is categorical; hence, each predictor variable is cross-tabulated with the outcome variable, and the Pearson Chi-square test was performed to identify the association between each independent variable and the outcome variable. Pearson Chi-square test is a robust non-parametric statistical method appropriate for analyzing nominal or ordinal data. It is instrumental when dealing with unequal sample sizes and non-normal data distributions. This test provides valuable insights into the relationship between variables (McHugh, 2013). A multinomial logistic regression method predicted the association between commute mode and demographic, socioeconomic and neighbourhood environmental characteristics. The commute mode choice is categorized into four levels: independent walking, walking with an accompanying person, cycling and commuting on a motorized vehicle. Odds ratios (OR) are estimated in the model with motorized vehicular commutes as a reference category. Table 2 shows the variables used in this study.

Table 2. Variables used in the study

Measure	Scale or response category (and assigned code for analysis)
Demographic and Socioeconomic characteristics
Sex of the Child	1 = Girl, 2 = Boy.
Education of the Child (grade)	Exact Entered Number
Number of Children in the Home	Exact Entered Number
Occupation of the Parent	1 = Both are Working, 2 = Single Parent Working, 3 = Both are not Working.
Annual Income of the Family (in Indian Rupees)	1 = Below 3,00,000; 2 = 3,00,000 - 6,00,000; 3 = 6,00,000 - 12,00,000; 4 = Above 12,00,000.
Education of the Parent	1 = did not attend the schooling, 2 = Matriculation, 3 = Graduation, 4 = Above graduation.
School commute
Distance of the school from home	1 = <1km, 2 = 1 to 2km, 3 = 2-5km, 4 = above 5km.
Mode of school commute	1 = walking independently, 2 = walking with accompanying person, 3 = cycling, 4 = driven by parent/caregiver, 5 = commuting on auto/rickshaw/van/school bus, etc., 6 = public transport.
average school commute time in a day	1 = < 30minutes, 2 = 30-60 minutes, 3 = 60-120 minutes, 4 = >120 minutes.
Types of roads in the home-to-school route	1= minor and major streets, 2 = streets and major roads.
Traffic claiming measures installed at major intersections	1 = No, 2 = Yes.
Availability of designated cycle tracks	= No, 2 = Yes.

Figure 4. School sites and road types in selected study areas.

Results

Sample Characteristics

Among the overall 792 responses, inner metropolitan consists of n = 193 (24.3%), suburban established consists of n = 211 (26.64%), suburban isolated consists of n = 220 (27.7%), and transient neighbourhood consists of n = 168 (21.21%) responses. Table 3 provides an overview of the sample and its demographic, socioeconomic, home-to-school distance, and commute mode characteristics with respect to its neighbourhoods.

Table 3. Characteristics of the sample across four neighbourhoods

Variable	Inner metropolitan	Suburban established	Suburban isolated	Transient	Overall
	Frequency (%)	Frequency (%)	Frequency (%)	Frequency (%)	Frequency (%)
Number of Respondents	193 (24.36)	211 (26.64)	220 (27.7)	168 (21.3)	792 (100)
Sex of the Child
Male	87 (45.1)	119 (56.4)	127 (57.7)	91 (54.2)	424 (53.5)
Female	106 (54.9)	92 (43.6)	93 (42.3)	77 (45.8)	368 (46.5)
Level of Education of the Child
Primary	105 (54.4)	98 (46.4)	101 (45.9)	88 (52.4)	392 (49.5)
High School	88 (45.6)	113 (53.6)	119 (54.1)	80 (47.6)	400 (50.5)
Number of Children in the Home
1	100 (51.8)	113 (53.6)	115 (52.3)	87 (51.8)	415 (52.4)
2	92 (47.7)	96 (45.5)	96 (43.6)	73 (43.5)	357 (45.1)
3	1 (0.5)	2 (0.9)	9 (4.1)	8 (4.8)	20 (2.5)
Education of the parent
Did not attend the school	2 (1)	2 (0.9)	9 (4.1)	21 (12.5)	34 (4.3)
Up to Matriculation	5 (2.6)	7 (3.3)	24 (10.9)	46 (27.4)	82 (10.4)
Graduate	92 (47.7)	107 (50.7)	109 (49.5)	62 (36.9)	370 (46.7)
Above Graduation	94 (48.7)	95 (45)	78 (35.5)	39 (23.2)	306 (38.6)
Occupation of the parent
Both are not working	2 (1)	2 (0.9)	8 (3.6)	25 (14.9)	37 (4.7)
Single parent working	140 (72.5)	162 (76.8)	153 (69.5)	106 (63.1)	561 (70.8)
Both are working	51 (26.4)	47 (22.3)	59 (26.8)	37 (22)	194 (24.5)
Annual Income of the family (Indian Rupees)
Below 3,00,000	26 (13.5)	40 (19)	60 (27.3)	89 (53)	215 (27.1)
3,00,001 – 6,00,000	61 (31.5)	129 (61.1)	95 (43.2)	58 (34.5)	343 (43.3)
6,00,001 – 12, 00,000	79 (41)	28 (13.3)	46 (21)	18 (10.7)	171 (21.6)
Above 12,00,000	27 (14)	14 (6.6)	19 (8.5)	3 (1.8)	63 (8)
Distance of the School
< 1 km	46 (23.8)	75 (35.5)	50 (22.7)	53 (31.5)	224 (28.3)
1-2 km	47 (21.5)	67 (31.8)	49 (22.3)	56 (33.3)	219 (27.7)
2-5 km	64 (33.2)	69 (32.7)	79 (35.9)	52 (31)	264 (33.3)
> 5 km	36 (18.7)	-	42 (19.1)	7 (4.2)	85(10.7)
Mode of commute
Independently Walking	29 (15)	53 (25.1)	39 (17.7)	39 (23.2)	160 (20.2)
Walking by an accompanying person	24 (12.4)	28 (13.3)	33 (15)	27 (16.1)	112(14.1)
Cycling	29 (15)	48 (22.7)	22 (10)	23 (13.7)	122 (15.4)
Private Travel (Family vehicle, Auto/Rickshaw, School Van/Bus, etc.)	107 (55.4)	82(38.9)	109 (49.5)	58 (34.5)	356 (44.9)
Public Transport	4 (2.1)	0	17 (7.7)	21 (12.5)	42 (5.3)
Route Type
Streets & minor roads	43 (22.27)	148 (70.1)	45 (20.5)	55 (32.7)	261 (32.95)
Major roads	150 (77.73)	93 (29.9)	175 (79.5)	113 (67.3)	531 (67.05)

Proportion of Children Studying Within and Above 1km Distance From Home

Of all the neighbourhoods, the proportion of children in suburban established (35.5%) and transient (31%) neighbourhoods attending school within 1km is relatively higher than the children in inner metropolitan (23.8%) and suburban isolated (22.7%) neighbourhoods. It indicates that the distance of the school varies with the neighbourhood characteristics, as parents in some neighbourhoods show a stronger preference for local schools while others prefer schools outside the neighbourhood. Table 4 and Table 5 represents the proportion of children attending school within 1km and above 1km from home.

Table 4. Neighbourhood type and home-to-school distance within 1km vs above 1km

Child's school distance from home ( < 1k vs > 1km)	Transient	Suburban isolated	Inner metropolitan	Suburban established	Overall
	Frequency (%)	Frequency (%)	Frequency (%)	Frequency (%)	Frequency (%)
Number of Respondents	168	220	193	211	792
Within 1km distance from home	53 (31)	50 (22.7)	46 (23.8)	75 (35.5)	224 (28.2)
More than 1km distance from home	115 (69)	170 (77.3)	147 (76.2)	136 (64.5)	568 (71.8)

Table 5. Travel mode share in different distances across all neighbourhoods

Distance	Commute Mode	Inner Metropolitan Freq. (%)	Suburban Established Freq. (%)	Suburban Isolated Freq. (%)	Transient Freq. (%)
< 1 km	Independently walking	25 (13)	44 (20.9)	28 (12.7)	29 (17.3)
	Walking with an accompanying person	19 (9.8)	23 (10.9)	19 (8.6)	22 (13.1)
	Cycling	2 (1)	5 (2.4)	2 (0.9)	2 (1.2)
	Commuting on a motorised vehicle	-	3 (1.4)	1 (0.5)	-
1-2 km	Independently walking	5 (2.6)	9 (4.3)	11 (5)	10 (6)
	Walking with an accompanying person	6 (3.1)	5 (2.4)	14 (6.4)	10 (6)
	Cycling	10 (5.2)	23 (10.9)	10 (4.5)	9 (5.4)
	Commuting on a motorised vehicle	26 (13.5)	30 (14.2)	14 (6.4)	27 (16.1)
2-5 km	Independently walking	-	-	-	3 (1.8)
	Walking with an accompanying person	-	-	-	9 (5.4)
	Cycling	17 (8.8)	20 (9.5)	10 (4.5)	12 (7.1)
	Commuting on a motorised vehicle	50 (25.9)	49 (23.2)	69 (31.4)	28 (16.7)
> 5 km	Independently walking	-	-	-	-
	Walking with an accompanying person	-	-	-	-
	Cycling	-	-	-	-
	Commuting on a motorised vehicle	33 (17.1)	-	42 (19.1)	7 (4.2)

Correlation Among The Variables

Commute Mode Vs Neighbourhood

Of all 792 responses, 379 (47.85%) respondents reported motorized vehicular commute, 164 (20.7%) independently walking, 127 (16.03%) walking with accompanying person and 122 (15.4%) cycling as a school commute mode for their children. Our sample shows that motorized vehicles are the dominant mode of transportation across all neighbourhoods compared to walking and cycling. In the inner metropolitan, motorized vehicular commutes comprise 56.5%, independent walking 15.5%, cycling 15%, and walking with accompanying persons 13%, respectively. Similarly, in suburban established neighbourhoods, the motorized vehicular commute is 38.9%, followed by independent walking (25.1%), cycling (22.8%), and walking with an accompanying person (13.3%). Among all neighbourhoods, independent walking has seen a relatively higher proportion in suburban established and transient areas and lower in suburban isolated and inner metropolitan neighbourhoods. The chi-square test of independence was performed to examine the association between a child's commute mode and neighbourhood type. The relation between these variables was significant, χ² (12, N=792) = 82.08, p < 0.001.

Home-to-school Distance Vs Neighbourhood

Our overall sample represents at least 28.28% of children attending school within 1km, 56% within 2km and 89.6% (including 1km&2km) within 5km from their residence. The rest, 10.4% of children, travel above 5km for schooling. In association with the Neighbourhood, a significant proportion of children in suburban established (67%) and transient (65%) neighbourhoods attend schools closer to their residence than in suburban isolated (45%) and inner metropolitan (48.5%) neighbourhoods. Surprisingly, no response was received on the above 5km category in the suburban established neighbourhood. The chi-square independent test examined the association between the neighbourhood type and home-to-school distance. The relation between neighbourhood type and home-to-school distance is significant, χ² (9, N=792) = 68.20, p < 0.001 (Table 6).

Table 6. Chi-square (χ2) correlations of neighbourhood types vs commute mode, distance and time.

Variable	Inner metropolitan Freq. (%)	Suburban established Freq. (%)	Suburban isolated Freq. (%)	Transient Freq. (%)	χ2	df	p
Number of Respondents	193	211	220	168
Mode of Commute					82.08	12	< 0.001
Independently Walking	30 (15.5)	53 (25)	39 (17.7)	42 (25)
Walking with an accompanying person	25 (13)	28 (13.2)	33 (15)	41 (24.4)
Cycling	29 (15)	48 (22.7)	22 (10)	23 (13.7)
Private Travel (Family vehicle, School Van/Bus, etc.)	105 (54.5)	82 (39)	109 (49.5)	42 (25)
Public Transport	4 (2)	-	17 (7.7)	20 (11.9)
Distance of the School from Home					68.20	9	< 0.001
< 1 km	46 (23.8)	75 (35.5)	50 (22.7)	53 (31.5)
1 -2 km	47 (24.3)	67 (31.5)	49 (22.3)	56 (33.3)
2 – 5 km	67 (34.7)	69 (32.7)	79 (36)	52 (31)
> 5km	33 (17)	-	42 (19)	7 (4.2)
Average commute time to the school					86.50	6	< 0.001
< 30 minutes	119 (61.6)	190 (90)	112 (51)	101 (60)
30-60 minutes	53 (27.4)	21 (10)	83 (37.7)	56 (33.5)
> 60 minutes	21 (11)	-	25 ()11.3	11 (6.5)

Predictors of School Commute Mode

We conducted a multinomial logistic regression model to predict the odds of the children's school commute mode (dependent variable: commute mode choice 1 = independently walking, 2 = walking with accompanying person, 3 = cycling and 4 = motorised vehicular commute. Independent / predictor variables are the child's sex, education level, parent's education, occupation and family's annual income, home-to-school distance, route type and neighbourhood type). Table 7, Table 8, and Table 9 shows the predicted statistics of the model concerning the child's commute mode choice. The model significantly predicted the commute mode preference of independently walking, walking with an accompanying person and cycling versus motorised vehicular commute.

In the overall sample, the number of children in the home does not exhibit statistically significant associations with independent walking, walking with an accompanying person, or cycling to school. However, neighborhood-specific analysis reveals interesting patterns. In the Inner metropolitan area, the number of children becomes a significant predictor for walking with an accompanying person, with lower odds, contrasting with the Transient neighbourhood, where more children are associated with independent walking. Cycling to school is not consistently influenced by the number of children across neighbourhoods. The annual family income is a critical determinant in shaping commuting behaviours in the overall sample, with lower income associated with increased odds of active commuting modes. However, within specific neighbourhoods, the relationship between income and commuting behaviours varies, with the inner metropolitan and suburban established neighbourhoods showing different patterns compared to the suburban isolated and Transient neighbourhoods.

Sex is a significant determinant in commuting behaviours to school in the overall sample, with girls showing significantly lower odds of independent walking than boys. However, sex's impact varies across neighbourhoods. In the Inner metropolitan area, sex is not statistically significant for commute modes. In contrast, in the Suburban Isolated and Transient neighbourhoods, sex exhibits varied associations with independent walking and cycling. The child's education level emerges as a significant factor in determining commuting behaviours in the overall sample, with higher education associated with increased odds of independent walking and cycling. Neighbourhood-specific analysis reveals divergent patterns. The suburban established, suburban isolated and transient neighbourhoods show the significance of independent walking. The inner metropolitan area shows no significant association.

The distance to school is a pivotal factor shaping commuting behaviours, with a shorter distance associated with increased odds of active commuting modes. However, the relationship varies across neighbourhoods, highlighting the importance of proximity in promoting active commuting modes in urban and suburban settings. The findings underscore the universal significance of shorter distances, emphasizing the need for tailored interventions considering each neighbourhood's overall trend and specific characteristics. Road type significantly influences commuting behaviours, with streets and minor roads encouraging independent walking and walking with an accompanying person over major traffic roads. The impact varies across neighbourhoods, with the Inner metropolitan area and Suburban Established neighbourhood showing significant associations, while the Suburban Isolated neighbourhood displays no significant relationship. The Transient neighbourhood exhibits a considerable association with independent walking.

Table 7. Multinomial logistic regression predicting school commute mode choice: Odds of independent walking to school (95% confidence interval) (Motorized vehicular commute is reference category)

Characteristic	Overall Sample a	Inner-metropolitan b	Suburban Established c	Suburban Isolated d	Transient e
Number of children in the home	0.754 (0.48-1.68)	0.251 (0.024-2.66)	0.814 (0.17-3.73)	1.448 (0.346-6.06)	0.627 (0.21-1.81)
Sex of the child (male = ref)	0.313** (0.15-0.64)	1.435 (0.159-12.94)	0.503 (0.10-2.46)	0.541 (0.08-3.32)	0.194* (0.05-0.70)
Level of education of the child (low to high)	3.28** (2.53-4.25)	8.737** (3.58-21.27)	5.16** (2.91-9.17)	4.347** (2.11-8.93)	2.265** (1.42-3.60)
Annual income of the family (low to high)	0.15** (0.084-0.27))	0.042** (0.007-0.252)	0.055** (0.01-0.22)	0.131* (0.03-0.53)	0.412 (0.106-1.6)
Education of the parent (low to high)	0.61 (0.66-2.01)	0.37 (0.37-25.82)	1.85 (0.35-9.84)	0.81 (0.21-3.07)	0.531 (0.20-1.41)
Distance to school (low to high)	0.004** (0.002-0.009)	0.001**	0.001**	0.001**	0.058** (0.02-0.16)
Road type (major roads = ref)
Streets & minor roads	4.67** (2.104-10.405)	1.19 (0.079-18.31)	1.83 (0.375-8.95)	16.54* (1.53-178.1)	6.58* (1.31-33.0)
Neighbourhood (Suburban established = ref)
Inner metropolitan	3.12* (1.103-8.87)	-	-	-	-
Suburban isolated	1.559 (0.57-4.19)	-	-	-	-
Transient	1.725 (0.60-4.93)	-	-	-	-

Note: ^a – Chi-square = 1182.29 (df = 30), main effects, sig = <0.001, R² = 0.592 (Mc Fadden), 0.843 (Nagelkerke), 0.775 (Cox and Snell), 78% correctly predicted. ^b – Chi-square = 331.707 (df = 21), main effects, sig = <0.001, R² = 0.740 (Mc Fadden), 0.910 (Nagelkerke), 0.821 (Cox and Snell), 88.6% correctly predicted. ^c – Chi-square = 361.39 (df = 21), main effects, sig = <0.001, R² = 0.649 (Mc Fadden), 0.883 (Nagelkerke), 0.820 (Cox and Snell), 79.1% correctly predicted. ^d – Chi-square = 353.24 (df = 21), main effects, sig = <0.001, R² = 0.704 (Mc Fadden), 0.890 (Nagelkerke), 0.799 (Cox and Snell), 87.3% correctly predicted. ^e – Chi-square = 228.03 (df = 21), main effects, sig = <0.001, R² = 0.510 (Mc Fadden), 0.798 (Nagelkerke), 0.743 (Cox and Snell), 69.6% correctly predicted.

Table 8. Multinomial logistic regression predicting school commute mode choice: Odds of walking with accompanying person to school (95% confidence interval) (Motorized vehicular commute is reference category)

Characteristic	Overall Sample a	Inner-metropolitan b	Suburban Established c	Suburban Isolated d	Transient e
Number of children in the home	0.547* (0.302-0.99)	0.077* (0.008-0.765)	0.887 (0.21-3.66)	0.489 (0.13-1.85)	0.834 (0.29-2.36)
Sex of the child (male = ref)	0.765 (0.402-1.45)	1.142 (0.15-8.69)	1.86 (0.35-9.92)	3.139 (0.6-16.31)	0.272* (0.07-1.02)
Level of education of the child (low to high)	0.650** (0.508-0.83)	1.341 (0.63-2.84)	0.908 (0.51-1.60)	0.761 (0.43-1.32)	0.288** (0.14-0.59)
Annual income of the family (low to high)	0.289** (0.17-0.487)	0.16* (0.03-0.84)	0.13* (0.03-0.53)	0.335 (0.11-1.03)	0.107* (0.02-0.45)
Education of the parent (low to high)	1.87* (1.068-3.30)	1.89 (0.360-9.99)	4.67 (0.84-25.9)	2.04 (0.60-6.88)	1.827 (0.62-5.33)
Distance to school (low to high)	0.033** (0.017-0.06)	0.001**	0.14** (0.003-0.07)	0.004** (0.00-0.043	0.160** (0.05-0.46)
Road type (major roads = ref)
Streets & minor roads	2.01* (1.02-3.95)	1.28 (0.12-13.03)	0.511 (0.11-2.34)	6.78 (0.90-51.0)	1.961 (0.54-7.06)
Neighbourhood (Suburban established = ref)
Inner metropolitan	1.99 (0.74-5.35)	-	-	-	-
Suburban isolated	3.22* (1.27-8.15)	-	-	-	-
Transient	4.419* (1.69-11.5)	-	-	-	-

Note: ^a – Chi-square = 1182.29 (df = 30), main effects, sig = <0.001, R² = 0.592 (Mc Fadden), 0.843 (Nagelkerke), 0.775 (Cox and Snell), 78% correctly predicted. ^b – Chi-square = 331.707 (df = 21), main effects, sig = <0.001, R² = 0.740 (Mc Fadden), 0.910 (Nagelkerke), 0.821 (Cox and Snell), 88.6% correctly predicted. ^c – Chi-square = 361.39 (df = 21), main effects, sig = <0.001, R² = 0.649 (Mc Fadden), 0.883 (Nagelkerke), 0.820 (Cox and Snell), 79.1% correctly predicted. ^d – Chi-square = 353.24 (df = 21), main effects, sig = <0.001, R² = 0.704 (Mc Fadden), 0.890 (Nagelkerke), 0.799 (Cox and Snell), 87.3% correctly predicted. ^e – Chi-square = 228.03 (df = 21), main effects, sig = <0.001, R² = 0.510 (Mc Fadden), 0.798 (Nagelkerke), 0.743 (Cox and Snell), 69.6% correctly predicted.

Table 9. Multinomial logistic regression predicting school commute mode choice: Odds of cycling to school (95% confidence interval) (Motorized vehicular commute is reference category)

Characteristic	Overall Sample a	Inner-metropolitan b	Suburban Established c	Suburban Isolated d	Transient e
Number of children in the home	1.095 (0.64-1.87)	3.02 (0.66-13.7)	0.235* (0.07-0.79)	2.304 (0.62-8.55)	1.225 (0.37-4.0)
Sex of the child (male = ref)	0.217** (0.11-0.40)	0.489 (0.116-2.07)	0.357 (0.10-1.21)	0.051* (0.01-0.33)	0.195* (0.04-0.8)
Level of education of the child (low to high)	3.471** (2.72-4.42)	5.93** (2.77-12.70)	4.087** (2.52-6.59)	6.42** (3.02-13.6)	3.516** (1.94-6.3)
Annual income of the family (low to high)	0.278** (0.17-0.45)	0.098* (0.02-0.41)	0.338* (0.13-0.85)	0.07** (0.01-0.31)	0.392 (0.11-1.3)
Education of the parent (low to high)	1.188 (0.76-1.85)	1.91 (0.47-7.61)	1.149 (0.40-3.25)	2.392 (0.71-8.01)	0.750 (0.28-1.95)
Distance to school (low to high)	0.069** (0.03-0.12)	0.005** (0.001-0.06)	0.057** (0.015-0.21)	0.004** (0.0-0.046)	0.256* (0.09-0.70)
Road type (major roads = ref)
Streets & minor roads	3.02* (1.338-6.83)	1.17 (0.101-13.5)	2.039 (0.59-6.96)	13.18 (0.86-202.1)	1.739 (0.13-22.7)
Neighbourhood (Suburban established = ref)
Inner metropolitan	1.82 (0.76-4.399)	-	-	-	-
Suburban isolated	0.35* (0.148-0.84)	-	-	-	-
Transient	0.749 (0.301-1.86)	-	-	-	-

Note: ^a – Chi-square = 1182.29 (df = 30), main effects, sig = <0.001, R² = 0.592 (Mc Fadden), 0.843 (Nagelkerke), 0.775 (Cox and Snell), 78% correctly predicted. ^b – Chi-square = 331.707 (df = 21), main effects, sig = <0.001, R² = 0.740 (Mc Fadden), 0.910 (Nagelkerke), 0.821 (Cox and Snell), 88.6% correctly predicted. ^c – Chi-square = 361.39 (df = 21), main effects, sig = <0.001, R² = 0.649 (Mc Fadden), 0.883 (Nagelkerke), 0.820 (Cox and Snell), 79.1% correctly predicted. ^d – Chi-square = 353.24 (df = 21), main effects, sig = <0.001, R² = 0.704 (Mc Fadden), 0.890 (Nagelkerke), 0.799 (Cox and Snell), 87.3% correctly predicted. ^e – Chi-square = 228.03 (df = 21), main effects, sig = <0.001, R² = 0.510 (Mc Fadden), 0.798 (Nagelkerke), 0.743 (Cox and Snell), 69.6% correctly predicted.

Discussion

This study examined how varied neighbourhood settings can impact children's school commuting behaviour in the Indian urban context. Earlier studies by (Singh, N. and Vasudevan, 2018; Tetali, Edwards, Roberts, 2016) on children's school commute behaviour in India have examined the association of commute mode based on demographic, socioeconomic and home-to-school distance factors. This study is unique in that apart from demographic, socioeconomic, and home-to-school distance factors, it allowed a comparison of distinct urban neighbourhood environments in children's school commute behaviour assessment.

In developed countries, many efforts were made at the policy and implementation level related to enhancing and improving school commutes, specifically active and independent mobility. For instance, the Safe Routes to School (SRTS) programs in the US (Chriqui, Taber et al., 2012), School Travel Plans (STPs) in Canada (Wong, B. Y.-M., Faulkner et al., 2011), Walking School Bus (WSB) programs in the UK, US and New Zealand (Carlson, Steel et al., 2020), Bike-to-school initiatives in the Netherlands, traffic-claiming measures and Car-free zones around schools in US, UK, Canada and parts of Europe and Asia. While fundamental disparities in realities exist between developed countries and India, their practices still imply the need for systematic research. Such research could enhance our current comprehension of children's school travel mode choices and contribute valuable insights to developing future interventions to create an improved travel environment for children in urban areas in India.

Firstly, urbanization in India is different from that in Western countries. In the context of this study, Visakhapatnam is a historic port city in India. The city is surrounded by the Bay of Bengal Ocean on the east and Eastern Ghats on the west. The past growth of the city radiated from the port in all three directions. However, the recent growth has been noticed in a linear expansion along the National Highway-16 in the south and north directions (WorldBank, 2012) And the four neighbourhood areas studied are distinct in their urbanization characteristics (See Table 1). The findings of our study indicate that home-to-school distance has a consistent and significant association with commute mode choice across all the neighbourhoods. The results show approximately 30% of the children attend schools within 1km and 90% in 5km, and only 10% are commuting beyond 5km from their residence. This finding is similar to earlier studies in various contexts(Broberg and Sarjala, 2015; Mehdizadeh, Mamdoohi et al., 2017; Mitra, Buliung et al., 2010). In previous studies, the maximum threshold for walking distances for school commutes in other countries was reported to be 1.5 km in Canada (Mammen, Stone et al., 2014), 2 km in Ireland (Kelly and Fu, 2014) 3 km in China (Li and Zhao, 2015) and Netherlands (Goeverden and Boer, 2013). In Indian cities, 3 km in Hyderabad (Tetali, Edwards, Roberts, 2016) and 2 km in Kanpur (Singh, N. and Vasudevan, 2018). In line with the above studies, our study finds that independent walking and accompanying was seen up to 2km in inner metropolitan, suburban established and isolated areas. However, in the transient neighbourhood, independent walking and walking with accompanying has been observed above the 2 km range. These findings are consistent in predominantly urban areas and differ in transient areas due to differences in the spatial distribution of schools.

Regarding cycling, suburban established neighbourhoods display a moderate prevalence, particularly in the 2-5 km range, suggesting a preference for this mode in slightly longer distances. Suburban isolated areas show a varied but consistent use of bicycles, particularly for medium distances, indicating a potential inclination toward cycling as a commuting option. In transient neighbourhood, cycling is more irregular, with significant usage observed only for distances between 2-5 km. Nearly half of the children across all neighbourhoods commute on motorized vehicles. Especially in the 2-5 km and > 5 km distance categories, with the highest frequencies observed in the suburban established and transient neighbourhoods. It underscores a consistent reliance on motorized transportation for relatively longer distances. According to the (Staitsta, 2015) report, the household ownership of a car in India stands at 6%, much lower to countries like the US 88%, Europe 85%, Japan 81%, UK 74%, and Asian countries Thailand 51% and China 17%. It emphasizes that autos, school buses, two-wheelers, and public transportation are dominant motorized vehicular school commutes in Indian cities. The data implies that the choice of motorized vehicles for school commutes is influenced by the distance to school and the specific characteristics of the neighbourhoods, highlighting the need for targeted interventions to address commuting behaviours in each context.

Regarding the child's characteristics, sex emerges as a crucial determinant in commuting behaviours, with boys exhibiting higher odds of independent walking. It aligns with the established findings of (Panter, Jones et al., 2010), documenting sex differences in active transportation patterns among children. The subtle associations observed across neighbourhoods underscore the need for targeted interventions that address sex-specific factors within diverse urban settings. However, our modelling results found that the child's education level is a crucial determinant of commute mode rather than sex itself; as the level of education and sex of the child together influence commute mode, these differences are highly significant in suburban isolated and transient neighbourhoods.

While considering the household factors, the impact of annual family income on commuting behaviours aligns with a complex relationship between socioeconomic factors and active transportation. Suburban established, and inner metropolitan areas are equipped with a variety of schools over the other two neighbourhoods. Parents of high-income choose better schools irrespective of distance constraints in inner metropolitan and suburban isolated areas; in both areas, more than 17 % of children travel above 5 km for schooling, and 100 % of those commutes are by motorised vehicles. The variations across neighbourhoods underscore the intricate interplay of income and local context, aligning with (Mitra and Buliung, 2015) recommendations for socio-economically sensitive interventions in urban and suburban settings.

Route characteristics emerge as a significant influencer, aligning with (Panter, Jones et al., 2010) study emphasizing the role of road infrastructure in shaping commuting behaviours. Our study's varying impact across neighbourhoods underscores the need for context-specific interventions, acknowledging the diverse road environments that children navigate in urban and suburban contexts. Research on the safety implications of different road types highlights the need for comprehensive road infrastructure planning to promote safe and active transportation for school children.

Regarding safety, each alternative mode of travel possesses its distinct competitive range. For example, all the school buses in India follow strict safety measures like providing additional caretakers in the school bus to monitor the child's pick-up and drop-off at designated locations and handing them over to the concerned parent or caretaker. It is also a prevalent practice similar to alternative private transports like auto and rickshaw modes. This enhances the confidence of parents of primary school children in assigning school buses over allowing independent walking and escorting. Also, our statistical results suggest optimal scope for cycling up to 5 km distance. Therefore, attention should be directed towards infrastructure that supports cycling. In India, the exponential growth of automobiles and improper cycling-friendly road networks are causing a threat to the disappearance of bicycle use in many urban areas. Urban areas should be devoted to encouraging and promoting cycling within the optimal distance, with separate bicycle lanes and supporting services.

Based on descriptive and modeling results from this study, appropriate facilities should be designed to support functioning and maximize active commute modes. As shown in Table 5 (distribution of commute mode) and from Table 7 to Table 9 (odds of commuting mode), walking to school ranges from 0-2 km. Within this, a significant proportion is relying on motorized vehicles. Therefore, the policies and urban planners should focus on creating a pedestrian-friendly neighbourhood environment to promote walkability in urban areas. On the other hand, school timing is one of the significant factors in escorting behaviour. In Visakhapatnam, different schools follow different timings for different grades. Thus, it is convenient for some parents to drop off and pick up their children in route to their workplaces.

In summary, the results of this study contribute to the evolving discourse on children's commuting behaviours, emphasizing the need for dynamic and context-specific interventions. Recent studies have shown a growing recognition of the importance of specific strategies that consider both individual and contextual factors. The field is increasingly focusing on distinctive analyses addressing the unique challenges of urban and suburban landscapes, highlighting the importance of interdisciplinary and multi-dimensional approaches to promoting sustainable and active transportation options for school-commuting children.

Conclusion

This study is an effort to identify the school commuting behaviour in urban India, particularly in various neighbourhood contexts. This study is unique in considering spatio-temporal characteristics of diverse urban neighbourhoods. This study used a cross-sectional survey and multinomial logistic regression in modelling commute behaviour. This method was widely used in earlier studies in different contexts (Goeverden and Boer, 2013; Mehdizadeh, Mamdoohi et al., 2017; Mitra and Buliung, 2015; Singh, N. and Vasudevan, 2018; Tetali, Edwards, Roberts, 2016).

The findings of this study underscore the consistent significance of home-to-school distance in shaping commute mode choices across all neighbourhoods. The study identifies a prevalent pattern of independent walking and walking with accompaniment up to 2 km in inner metropolitan, suburban established, and suburban isolated areas. Conversely, transient neighbourhoods exhibit this pattern beyond the 2 km range, highlighting spatial distribution variations of schools. Cycling patterns reveal a moderate preference in suburban established areas for distances of 2-5 km, while suburban isolated areas display varied yet consistent bicycle use. Transient neighbourhoods show irregular cycling, emphasizing the need for context-specific interventions. Motorized vehicles dominate for longer distances, reflecting India's lower car ownership rates compared to developed countries. The study stresses the importance of targeted interventions addressing commuting behaviours in diverse urban settings.

Sex emerges as a crucial determinant, with boys exhibiting higher odds of independent walking. However, the child's level of education proves to be a key factor, with significant interactions between education and sex in suburban isolated and transient neighbourhoods. Household factors, precisely annual family income, influence commuting behaviours, with variations across neighbourhoods highlighting the intricate interplay of income and local context. Route characteristics and safety considerations also play a pivotal role, emphasizing the need for context-specific interventions and comprehensive road infrastructure planning. The study advocates for cycle-friendly infrastructure and separate bicycle lanes within optimal distances.

In conclusion, the study highlights the evolving discourse on sustainable and active transportation options, emphasizing the importance of interdisciplinary approaches. The findings provide valuable insights for policymakers and urban planners to focus on creating pedestrian-friendly neighbourhoods and interventions for the unique challenges posed by different urban and suburban landscapes. In this study, the escorting behaviour by walking was only segregated and not documented separately on two- and four-wheeler commutes.

Authors Contributions

Conceptualization, R.R.; methodology, R.R.; software, R.R.; investigation, R.R.; resources, R.R.; data curation, R.R.; writing—original draft preparation, R.R.; writing—review and editing, R.R. and R.S.P.; supervision, R.S.P. All authors have read and agreed to the published version of the manuscript.

Ethics Declaration

The authors declare that they have no conflicts of interest regarding the publication of the paper.

Funding

On behalf of all authors, the corresponding author states that this research received no external funding.

Acknowledgments

We thank all the parents participated in this study and the assistance provided by the local government authorities of Visakhapatnam City in secondary data collection.

References

•  Adlakha,  D.,  Hipp , J. A., et al. (2018). "Exploring Neighborhood Environments and Active Commuting in Chennai, India". International Journal of Environmental Research and Public Health, 15(9). doi: http://doi.org/10.3390/ijerph15091840.
•  Bi, B. and  Zhang,  X. (2016). "Attending Nearby Schools” in Central Beijing: Influencing Factors and the Policy Distortion". International Review for Spatial Planning and Sustainable Development, 4(1), 31-48. doi: http://doi.org/10.14246/irspsd.4.1_31.
•  Broberg,  A. and  Sarjala,  S. (2015). "School Travel Mode Choice and the Characteristics of the Urban Built Environment: The Case of Helsinki, Finland". Transport Policy, 37, 1-10. doi: https://doi.org/10.1016/j.tranpol.2014.10.011.
•  Carlson , J. A.,  Steel,  C., et al. (2020). "Walking School Bus Programs: Implementation Factors, Implementation Outcomes, and Student Outcomes, 2017-2018". Prev Chronic Dis, 17, E127. doi: http://doi.org/10.5888/pcd17.200061.
•  Carver,  A.,  Timperio,  A., et al. (2013). "Parental Chauffeurs: What Drives Their Transport Choice?". Journal of Transport Geography, 26, 72-77. doi: https://doi.org/10.1016/j.jtrangeo.2012.08.017.
•  Chriqui , J. F.,  Taber , D. R., et al. (2012). "The Impact of State Safe Routes to School-Related Laws on Active Travel to School Policies and Practices in U.S. Elementary Schools". Health & Place, 18(1), 8-15. doi: https://doi.org/10.1016/j.healthplace.2011.08.006.
•  Davison , K. K.,  Werder , J. L., et al. (2008). "Children's Active Commuting to School: Current Knowledge and Future Directions". Prev Chronic Dis, 5(3), A100.
•  Easton,  S. and  Ferrari,  E. (2015). "Children's Travel to School—the Interaction of Individual, Neighbourhood and School Factors". Transport Policy, 44, 9-18. doi: https://doi.org/10.1016/j.tranpol.2015.05.023.
•  Faulkner , G. E.,  Richichi,  V., et al. (2010). "What's "Quickest and Easiest?": Parental Decision Making About School Trip Mode". International Journal of Behavioral Nutrition and Physical Activity, 7(1), 62. doi: http://doi.org/10.1186/1479-5868-7-62.
•  Faulkner , G. E. J.,  Buliung , R. N., et al. (2009). "Active School Transport, Physical Activity Levels and Body Weight of Children and Youth: A Systematic Review". Preventive Medicine, 48(1), 3-8. doi: https://doi.org/10.1016/j.ypmed.2008.10.017.
•  Goeverden , C. D. v. and  Boer , E. d. (2013). "School Travel Behaviour in the Netherlands and Flanders". Transport Policy, 26, 73-84. doi: https://doi.org/10.1016/j.tranpol.2013.01.004.
• He, S. Y. (2013). "Will You Escort Your Child to School? The Effect of Spatial and Temporal Constraints of Parental Employment". Applied Geography, 42, 116-123. doi: https://doi.org/10.1016/j.apgeog.2013.05.003.
•  Hsu,  H.-P. and  Saphores,  J.-D. (2014). "Impacts of Parental Gender and Attitudes on Children’s School Travel Mode and Parental Chauffeuring Behavior: Results for California Based on the 2009 National Household Travel Survey". Transportation, 41(3), 543-565. doi: http://doi.org/10.1007/s11116-013-9500-7.
•  Kelly , J. A. and  Fu,  M. (2014). "Sustainable School Commuting – Understanding Choices and Identifying Opportunities: A Case Study in Dublin, Ireland". Journal of Transport Geography, 34, 221-230. doi: https://doi.org/10.1016/j.jtrangeo.2013.12.010.
•  Lee , J. S. (2020). "Urban Form, Children’s Active Travel to/from School, and Travel Related Physical Activity". International Review for Spatial Planning and Sustainable Development, 8(1), 21-38. doi: http://doi.org/10.14246/irspsd.8.1_21.
•  Li, S. and  Zhao,  P. (2015). "The Determinants of Commuting Mode Choice among School Children in Beijing". Journal of Transport Geography, 46, 112-121. doi: https://doi.org/10.1016/j.jtrangeo.2015.06.010.
•  Mammen,  G.,  Stone , M. R., et al. (2014). "School Travel Planning in Canada: Identifying Child, Family, and School-Level Characteristics Associated with Travel Mode Shift from Driving to Active School Travel". Journal of Transport & Health, 1(4), 288-294. doi: https://doi.org/10.1016/j.jth.2014.09.004.
•  Mandic,  S.,  Sandretto,  S., et al. (2023). "School Choice, Distance to School and Travel to School Patterns among Adolescents". Journal of Transport & Health, 33, 101704. doi: https://doi.org/10.1016/j.jth.2023.101704.
•  Mandic,  S.,  Williams,  J., et al. (2016). "Built Environment and Active Transport to School (Beats) Study: Protocol for a Cross-Sectional Study". BMJ Open, 6(5), e011196. doi: http://doi.org/10.1136/bmjopen-2016-011196.
•  Manjunatha,  Uthkarsh, P. S., et al. (2020). "Children Commute to Schools in Bangalore Urban and Rural Districts: Travel Pattern, Behaviour and Perceptions". International Journal Of Community Medicine And Public Health, 7(2), 748-755. doi: http://doi.org/10.18203/2394-6040.ijcmph20200461.
•  Marshall , J. D.,  Wilson , R. D., et al. (2010). "Vehicle Emissions During Children’s School Commuting: Impacts of Education Policy". Environmental Science & Technology, 44(5), 1537-1543. doi: http://doi.org/10.1021/es902932n.
•  Marzi,  I. and  Reimers , A. K. (2018). "Children’s Independent Mobility: Current Knowledge, Future Directions, and Public Health Implications". International Journal of Environmental Research and Public Health, 15(11), 2441.
•  McDonald , N. C. (2008). "Children’s Mode Choice for the School Trip: The Role of Distance and School Location in Walking to School". Transportation, 35(1), 23-35. doi: http://doi.org/10.1007/s11116-007-9135-7.
•  McHugh , M. L. (2013). "The Chi-Square Test of Independence". Biochemia Medica, 143-149. doi: https://doi.org/ 10.11613/bm.2013.018.
•  McMillan , T. E. (2005). "Urban Form and a Child’s Trip to School: The Current Literature and a Framework for Future Research". Journal of Planning Literature, 19(4), 440-456. doi: https://doi.org/10.1177/0885412204274173.
•  McMillan , T. E. (2007). "The Relative Influence of Urban Form on a Child’s Travel Mode to School". Transportation Research Part A: Policy and Practice, 41(1), 69-79. doi: https://doi.org/10.1016/j.tra.2006.05.011.
•  Mehdizadeh,  M.,  Mamdoohi,  A., et al. (2017). "Walking Time to School, Children's Active School Travel and Their Related Factors". Journal of Transport & Health, 6, 313-326. doi: https://doi.org/10.1016/j.jth.2017.01.012.
•  Mitra,  R. (2013). "Independent Mobility and Mode Choice for School Transportation: A Review and Framework for Future Research". Transport Reviews, 33(1), 21-43. doi: https://doi.org/10.1080/01441647.2012.743490.
•  Mitra,  R. and  Buliung , R. N. (2015). "Exploring Differences in School Travel Mode Choice Behaviour between Children and Youth". Transport Policy, 42, 4-11. doi: https://doi.org/10.1016/j.tranpol.2015.04.005.
•  Mitra,  R.,  Buliung , R. N., et al. (2010). "Built Environment and School Travel Mode Choice in Toronto, Canada". Transportation Research Record, 2156(1), 150-159. doi: https://doi.org/10.3141/2156-17.
• MORTH. (2023). "Road Transport Year Book (2019-20)". New Delhi: Ministry of Road Transport and Highways, Research Wing, Government of India. Retrieved from https://morth.gov.in/en/road-transport-year-books.
•  Omura , J. D.,  Hyde, E. T.,  et al. (2019). "Prevalence of Children Walking to School and Related Barriers—United States, 2017". Preventive Medicine, 118, 191-195. doi: https://doi.org/10.1016/j.ypmed.2018.10.016.
•  Panter , J. R.,  Jones , A. P., et al. (2010). "Neighborhood, Route, and School Environments and Children's Active Commuting". American Journal of Preventive Medicine, 38(3), 268-278. doi: https://doi.org/10.1016/j.amepre.2009.10.040.
•  PIB . (2022). "Growth in the Number of School in India". AKAM Series#2. Retrieved from https://www.pib.gov.in/FactsheetDetails.aspx?Id=148682 on 02-06-2025, 2025.
•  Reddy,  R. and  Pasupuleti , R. S. (2024). "Parental Preferences and Non-Spatial Factors Influencing Home-to-School Distance in Urban India: Insights from Visakhapatnam". Journal of Contemporary Urban Affairs, 8(1). doi: https://doi.org/ 10.25034/ijcua.2024.v8n1-13.
•  Seaton,  A.,  Godden,  D., et al. (1995). "Particulate Air Pollution and Acute Health Effects". The Lancet, 345(8943), 176-178. doi: https://doi.org/10.1016/S0140-6736(95)90173-6.
•  Sharmin,  S. and  Kamruzzaman,  M. (2017). "Association between the Built Environment and Children's Independent Mobility: A Meta-Analytic Review". Journal of Transport Geography, 61, 104-117. doi: https://doi.org/10.1016/j.jtrangeo.2017.04.004.
•  Sidharthan,  R.,  Bhat , C. R., et al. (2011). "Model for Children's School Travel Mode Choice:Accounting for Effects of Spatial and Social Interaction". Transportation Research Record, 2213(1), 78-86. doi: https://doi.org/10.3141/2213-11.
• Singh,  D.,  Singh, S. P., et al. (2016). "Epidemiology of Road Traffic Accident Deaths in Children in Chandigarh Zone of North West India". Egyptian Journal of Forensic Sciences, 6(3), 255-260. doi: https://doi.org/10.1016/j.ejfs.2015.01.008.
•  Singh,  N. and  Vasudevan,  V. (2018). "Understanding School Trip Mode Choice – the Case of Kanpur (India)". Journal of Transport Geography, 66, 283-290. doi: https://doi.org/10.1016/j.jtrangeo.2017.12.007.
•  Srinivasan,  M. (2010, 24-08-2010). How Traffic Congestion Affects School Students, Online The Hindu. Retrieved from https://www.thehindu.com/news/cities/chennai/How-traffic-congestion-affects-school-students/article16142542.ece
•  Staitsta . (2015). "Percentage of Households Owning a Car in Selected Countries in 2014, by Country". Retrieved from https://www.statista.com/statistics/516280/share-of-households-that-own-a-passenger-vehicle-by-country/ on 02-06-2025, 2025.
•  Sun,  W.,  Guo,  D., et al. (2021). "School Runs and Urban Traffic Congestion: Evidence from China". Regional Science and Urban Economics, 86, 103606. doi: https://doi.org/10.1016/j.regsciurbeco.2020.103606.
•  Sun,  Y.,  Liu,  Y., et al. (2015). "Associations between Active Commuting to School, Body Fat, and Mental Well-Being: Population-Based, Cross-Sectional Study in China". Journal of Adolescent Health, 57(6), 679-685. doi: https://doi.org/10.1016/j.jadohealth.2015.09.002.
•  Tetali,  S.,  Edwards,  P., et al. (2016). "Road Traffic Injuries to Children During the School Commute in Hyderabad, India: Cross-Sectional Survey". Injury Prevention, 22(3), 171-175. doi: https://doi.org/10.1136/injuryprev-2015-041854.
•  Tetali,  S.,  Edwards,  P., et al. (2016). "How Do Children Travel to School in Urban India? A Cross-Sectional Study of 5,842 Children in Hyderabad". BMC Public Health, 16(1), 1099. doi: https://doi.org/10.1186/s12889-016-3750-1.
•  Timperio,  A.,  Ball,  K., et al. (2006). "Personal, Family, Social, and Environmental Correlates of Active Commuting to School". American Journal of Preventive Medicine, 30(1), 45-51. doi: https://doi.org/10.1016/j.amepre.2005.08.047.
•  Waygood , E. O. D. and  Kitamura,  R. (2009). "Children in a Rail-Based Developed Area of Japan:Travel Patterns, Independence, and Exercise". Transportation Research Record, 2125(1), 36-43. doi: https://doi.org/10.3141/2125-05.
•  Wilson,  K.,  Coen , S. E., et al. (2019). "Children's Perspectives on Neighbourhood Barriers and Enablers to Active School Travel: A Participatory Mapping Study". Canadian Geographies / Géographies canadiennes, 63(1), 112-128. doi: https://doi.org/10.1111/cag.12488.
•  Wong B. Y.-M., Faulkner, G., et al. (2011). "Mode Shifting in School Travel Mode: Examining the Prevalence and Correlates of Active School Transport in Ontario, Canada". BMC Public Health, 11(1), 618. doi: https://doi.org/10.1186/1471-2458-11-618.
•  Wong,  C.,  Zheng,  W., et al. (2020). "Urban Expansion and Neighbourhood Commuting Patterns in the Beijing Metropolitan Region: A Multilevel Analysis". Urban Studies, 57(13), 2773-2793. doi: https://doi.org/10.1177/0042098019884254.
•  WorldBank . (2012). "City Development Plan : Greater Visakhapatnam". Washington, DC: World Bank. Retrieved from http://documents.worldbank.org/curated/en/195691468042841007.
•  Yang,  L.,  Wang,  B., et al. (2018). "Willing to Pay More for High-Quality Schools? A Hedonic Pricing and Propensity Score Matching Approach". International Review for Spatial Planning and Sustainable Development, 6(1), 45-62. doi: https://doi.org/10.14246/irspsd.6.1_45.