Fundamental Analysis of the Ages of Children and Road Structures Involved in Traffic Accidents

Abstract

The population of children in Japan has steadily declined, and the percentage of the population aged 14 years or younger is smaller than in other countries. Therefore, it is important to protect children by preventing their involvement in traffic accidents. Recent trends associated with elementary school students show that 90% of accidents occur while walking or riding bicycles. This study aims to clarify the road structures in which traffic accidents involving walking or bicycle-riding children occur. For this purpose, we analyzed the trends based on the age of children and road structures involved using statistical data provided by the Ishikawa Prefectural Police Headquarters. The results showed that the accident rate among students of elementary school age and younger while walking or riding bicycles was mainly high on one-lane roads, roads with no division, and roads with a speed limit of 30 km/h or less. We conclude that to reduce pedestrian and bicycle accidents for students of elementary school age or younger, raising traffic safety awareness among automobile drivers who use community roads is important.

1. Introduction

The population of children in Japan has been declining, and the percentage of the population aged 14 years or younger is 11.9%, which is smaller than the global percentage (25.4%) [1]. Furthermore, future projections estimate that the birth rate will continue to decline. This implies that protecting children is of paramount importance, and one method for achieving this objective is to prevent traffic accidents by maintaining a safe traffic environment and promoting traffic safety education. The 11th Fundamental Traffic Safety Plan formulated by the Central Traffic Safety Measures Council in March 2021 emphasizes “ensuring the safety of older adults and children” and “ensuring the safety of pedestrians and cyclists and improving awareness of legal compliance”, and it states that further measures to prevent traffic accidents are needed [2].

Examples of recent traffic safety initiatives for children include conducting emergency safety inspections on routes in which groups of children routinely travel, maintaining safe walking spaces, and implementing traffic safety measures on community roads, such as the introduction of Zone 30 Plus. The introduction of Zone 30 Plus is a measure to promote safe travel spaces that prioritize people on community roads by restricting the speed limit of vehicles to 30 km/h via physical methods (e.g., using humps, rising bollards, and smooth pedestrian crossings). Because of these efforts, traffic accidents involving elementary school students are declining [3]. However, in 2022, such traffic accidents caused 618 fatalities and severe injuries. Among them, 330 (approximately 54%) cases involved walking, and 235 cases (approximately 38%) involved riding bicycles, accounting for over 90% of the total incidents. Therefore, reducing accidents while walking or riding bicycles is important for ensuring the safety of elementary school students.

Several studies have focused on the traffic safety of pedestrians and cyclists. Saito analyzed the characteristics of traffic accidents in children aged up to 12 years and discussed measures for preventing characteristic accidents [4]. Although useful knowledge was obtained, the research was published more than 20 years ago, and the characteristics and influencing factors of traffic accidents, as well as road traffic conditions, have changed significantly in recent years. Miyazaki and Matsuo combined traffic accident and trip survey data to analyze the relationship between traffic accidents among walking children and traffic behavior [5]. An increased number of traffic accidents were observed during walking trips for children aged 5–7 years old. Miyazaki and Morimoto performed a comparative analysis of traffic accidents involving children on and off school routes in Utsunomiya City and found that most accidents on the school route occurred at low speeds [6]. Hagita analyzed traffic accidents and travel time for elementary school students on their way to and from their schools and other places [7]. The risk of traffic accidents was low while walking to and from school because fewer accidents occurred with respect to travel time, and the percentage of law violations while walking was low. Matsuo et al. verified a method to evaluate the risk of elementary school students–vehicle accidents considering the difference in travel purposes; they clarified that accident risk structure varies according to the purpose of the children’s travels [8]. Haga and Hamaoka analyzed the head-turning behavior of pedestrians to study the safety of pedestrian crossings [9]. Head-swinging increased toward the entry point of the crosswalk and at the crossing point. Deluka-Tibljaš et al. took up a study summary carried out in the world about the road safety of walking children at pedestrian crossings in the past 10 years [10]. For a further study, we suggested that the influence on road safety of the children was analyzed by distractions in detail. Campisi et al. investigated the distraction of pedestrians using mobile phone [11]. As a result, we showed that the danger of the accident increased because we tended to ignore the mark, and the walking speed in the road crossing decreased. Liu et al. investigated the effect that a distraction (sight and hearing) by the use of the mobile phone gave to the discrimination time of the pedestrian for the signal for the pedestrian [12]. As a result, the man 40 years or older made an announcement on being slow in reaction in comparison with other pedestrians. Sobhani and Farooq investigated the crossing action of the pedestrian, who was the state that was a distraction using a virtual reality environment [13]. As a result, the safety of the pedestrian who was a distraction was improved by establishing the LED right field where a color changed on the pedestrian crossing and showed that the crossing success rate was raised. Inagaki et al., who focused on pedestrian accidents involving children that occurred when they were crossing the street, conducted experiments to determine the effects of differences in vehicle distance and speed on the ability of elementary school children to make street-crossing judgments, as well as the timing of vehicle perception on children’s street-crossing judgments [14,15]. Their results showed that, compared to adults, elementary school children have more difficulty judging vehicle speed and that a shorter perceived distance induces misjudgments. Hyodo and Kobayashi analyzed the relationship between bicycle-related accidents and traffic exposure using data on bicycle and automobile traffic flow, traffic accidents, and intersection shapes [16]. Bicycle-related accidents and the traffic volume of automobiles and bicycles turning left or right from the main road to the secondary road were significantly correlated. To ascertain the distribution of bicycle-related accidents at intersections, Hagita et al. analyzed the frequency of accidents by traffic direction of the parties involved and showed that the presence or absence of a traffic signal resulted in different types of accidents involving bicycles and automobiles [17]. Terashima et al. examined the characteristics of traffic accidents involving bicycles with dual infant passengers using statistical data on traffic accidents [18]. Many accidents involved vehicles traveling at low speeds, and the major injury sites of cyclists and dual-infant passengers tended to differ. Toda and Ogura conducted crash tests of head-on collisions and discussed the importance of head protection in terms of differences in bicycle rider behaviors and head injury severity, depending on the impact site [19]. Kotake et al. conducted a questionnaire survey of students and teachers from elementary school to high school in Hyogo Prefecture, focusing on bicycle use, and observed that a lack of education, especially among junior and senior high school students, resulted in low interest in traffic safety and low awareness of violations [20]. Yano and Mori analyzed the number of young people, from first graders in elementary school through to third graders in high school, killed or injured in traffic accidents while walking or riding bicycles and categorized the numbers in terms of commuting to or from school, private use, and gender, focusing on changes after entering school [21]. Ciesla analyzed the safety level of children in pedestrian and bicycle traffic using accident data analysis and the investigation of satisfaction with the road infrastructure of road users [22]. She showed an intersection maintenance method that increases the safety of children. Xie et al. built a collision avoidance model and assessed the reduction in accidents by automated driving technology. Although the possibility of preventing accidents on community roads was confirmed, several accidents were difficult to prevent with vehicles alone [23].

Research studies have also focused on the relationship between road structure/environment and traffic safety. Hagita and Yokozeki focused on the road traffic environment when analyzing traffic accidents before and after earthquakes [24,25]. The rate of accidents in urban areas other than densely inhabited districts (DIDs) increased sharply, and the rate of accidents tended to be higher during the daytime than during the nighttime. Additionally, bicycle accidents on slope roads were more likely to occur on downhill roads, particularly bicycle-to-pedestrian and single-bicycle accidents. Hyodo et al. focused on human–vehicle accidents in non-arterial roads, creating a negative binomial regression model to analyze the relationship between accident risk and the characteristic factors of traffic/road/roadside considering the accident type [26]. Areas wherein the factors related to central business districts exhibited a higher trend tended to have a higher risk of front/rear and jaywalking accidents. Gürbüz and Buyruk developed “accident risk coefficients” to evaluate the accident risk and the model that decided the safe stopping distance of the vehicle based on a driver, a vehicle, and an environmental factor [27]. We clarified what we could use for the detection of the accident risk, that accident risk coefficients held in check to slight errors at speeds up to 60 km/h. Muramatsu et al. discussed the relationship between the walking characteristics of children and the school-commuting road environment [28]. The road traffic conditions that impact the crossing positions and cross-directional movement of children differed depending on the width of the school-commuting road. Shiomi et al. used Google Earth to quantify the distance between stop lines at intersections, road markings, and roadside land use and analyzed their relationship with accident risk using a Poisson regression model [29]. Making intersections more compact effectively reduced accident risk. Matsui et al. calculated the time to collision (TTC) by focusing on cases in which vehicles went straight, and bicycles crossed the road [30]. The TTC was significantly shorter when bicycles came out of the shadow of a building or vehicle than when they came from unobscured areas; therefore, head-on collisions were likely when obstacles reduced visibility. Suzuki et al. conducted a survey of the illuminance environment at locations where pedestrian accidents occurred at night and described trends associated with the illuminance environment that were not tabulated in statistical data on traffic accidents [31]. Yoshii et al. combined probe data and traffic accident data to conduct a factor analysis by building a multiple regression model [32]. The model employed the number of traffic accidents at community road intersections as the objective variable, and various factors as the explanatory variables, and the results clarified that the risk of accidents increased at four-way intersections with both poor and good visibility and at four-way intersections with narrow roads. Riaz et al. investigated if there are differences in crossing behavior related to road infrastructure, the gender of the child, and the effect of the accompanying adult on primary school children crossing behavior [33]. They showed that girls were about 1.9 times more likely to stop before crossing as compared to boys, and adults holding hands with the child resulted in safer behaviors by children. Congiu et al. matched with data on traffic accidents and spatial functional street qualities; on-street parking was found to increase the risk of pedestrian accidents [34].

The aforementioned literature indicates that numerous studies have been conducted to clarify the relationships between traffic safety for pedestrians and cyclists and road structure/environment and traffic safety. However, no studies have analyzed the characteristics of road structures that cause traffic accidents involving walking or bicycle-riding children, considering the age group.

Reducing traffic accidents while walking and riding bicycles is important for ensuring the traffic safety of children. Therefore, this study aims to clarify the road structures in which traffic accidents involving walking or bicycle-riding children are more likely to occur. In this regard, the traffic accident trends were analyzed according to the age of children and road structure using statistical data on traffic accidents obtained from the Ishikawa Prefectural Police Headquarters. The cases involving children up to junior high school level were analyzed, corresponding to an age range of 14 years or younger.

2. Materials and Methods

2.1. Traffic Accident Data Used

In this study, statistical data on traffic accidents recorded by the Ishikawa Prefectural Police Headquarters were used. Regarding the data, as stipulated in Article 2, Paragraph 1, Item 1 of the Road Traffic Act [35], traffic accidents are defined as accidents involving death or injury caused by a vehicle, streetcar, or train traffic on roads. For each case, information on the traffic accident and the parties involved is described as a single record. Among these cases, pedestrian-related (7952 cases) and bicycle-related (11,372 cases) accidents that occurred in the Ishikawa Prefecture from 2005 to 2020 were included in this study. In addition,

Table 1. Basic information of Ishikawa Prefecture.

Population:	1,132,526 (persons)
Area:	4186.21 km2
Population density:	270.5 person/km2
Population ratio of children:	12.1% (0–14 years old)

lists basic information about the target area: Ishikawa Prefecture [36].

The data items mainly included the date and time of the accident, attributes of the parties involved (type of party, sex, age, etc.), site information (latitude, longitude, road structure, etc.), and details of the accident (severity of injury, accident type, legal violation, etc.). For our analysis, the data for the type of party involved, the age of each party involved were used, and road structure (road shape, road width, sidewalk/road division, and speed limit).

Data pertaining to the type of party involved were recorded according to the following groups: automobile (passenger car, freight vehicle, and special vehicle), two-wheeled vehicle (motorcycle and motorized bicycle), bicycle, and pedestrian. From these data, the data on traffic accidents classified under “pedestrian” or “bicycle” were extracted.

The age of the children was subdivided into “6 years or younger (pre-schooler)”, “7–8 years (lower elementary school)”, “9–10 years (middle elementary school)”, “11–12 years (higher elementary school)”, and “13–15 years (junior high school)”.

Because the “road structure” item was not included in the data, we used the road shape, road width, sidewalk/road division, and speed limit as similar items. “Road shape” is divided into intersection, intersection vicinity, and single road; the analysis was explained according to the road shape in Section 3.1. “Road width” is divided into intersection and single road (includes intersection vicinity and railroad crossings). Intersections are divided based on the entry road of the first party involved as small (less than 5.5 m), medium (5.5 m or more), and large (13 m or more). In this study, we determined the number of lanes on the road from the road width using the lane width stipulated in the Road Construction Ordinance, assuming one lane as small, two lanes as medium, and four lanes or more as large. For single roads, the number of lanes was assumed in the same way. The analysis was explained according to the number of lanes in Section 3.2. For “sidewalk/road division”, four entries were available: division present (guard fence, curb/block, and roadside strip) and division absent categories. The analysis was explained according to the sidewalk/road division in Section 3.3. For “speed limit”, entries were for roads where the party involved was driving with a vehicle with two wheels or more, and 20–100 km/h are recorded every 10 km/h. The analysis was explained according to the speed limit in Section 3.4.

2.2. Analysis Method

We used the type and age of each party involved and the road structure (road shape, number of lanes, sidewalk/road division, and speed limit) to analyze the road structures for which traffic accidents involving children were common. For this objective, we calculated the number and percentage of pedestrian/cyclist-related accidents for all ages and age groups and then compared the accident percentage of each age group with that of all ages. We also judged significant differences using the Hypothesis Testing for the Difference in the Population Proportions.

Figure 1a,b show the other parties involved in pedestrian- and cyclist-related accidents. Over 95% of accidents involved automobiles; therefore, we tabulated the data and conducted analyses by limiting our cases to pedestrian–automobile accidents (henceforth referred to as “pedestrian accidents”; 7754 cases) and bicycle-automobile accidents (henceforth referred to as “bicycle accidents”; 11,035 cases).

3. Results and Discussion

3.1. Trends of Traffic Accidents Involving Children Depending on the Road Shape

Table 2. Percentage of pedestrian accidents according to the road shape and test results.

Age Division	Single Road Part		Intersection		Other	All Accidents (Cases)
	Percentage	p-Value *: p < 0.05 **: p < 0.01	Percentage	p-Value *: p < 0.05 **: p < 0.01	Percentage
Pre-schooler	56.96%	0.0000 **	22.03%	0.0000 **	21.01%	395
Lower elementary school	53.53%	0.0000 **	30.75%	0.0119 *	15.72%	439
Middle elementary school	53.62%	0.0002 **	30.92%	0.0887	15.46%	207
Higher elementary school	44.53%	0.3802	46.09%	0.0287 *	9.38%	128
Junior high school	42.95%	0.5695	44.23%	0.0532	12.82%	156
All ages	40.69%	1.0000	36.69%	1.0000	22.62%	7754

and

Table 3. Percentage of bicycle accidents according to the road shape and test results.

Age Division	Single Road Part		Intersection		Other	All Accidents (Cases)
	Percentage	p-Value *: p < 0.05 **: p < 0.01	Percentage	p-Value *: p < 0.05 **: p < 0.01	Percentage
Pre-schooler	25.00%	0.9361	52.78%	0.1257	22.22%	144
Lower elementary school	20.39%	0.0252 *	59.46%	0.8828	20.15%	407
Middle elementary school	20.91%	0.0478 *	64.23%	0.0406 *	14.86%	397
Higher elementary school	16.94%	0.0001 **	66.82%	0.0014 **	16.24%	431
Junior high school	23.24%	0.1186	62.28%	0.0321 *	14.47%	1209
All ages	25.29%	1.0000	59.09%	1.0000	15.61%	11,035

list the percentages of pedestrian and bicycle accidents with respect to the road shape, respectively, according to the age group. The “single road part” in the tables refers to single roads and intersection vicinities, and the same notation is used hereafter. Cases for which the percentage of accidents for an age group is higher than for all ages are shown in red hatching. Additionally, cases where the p-value is less than 0.05 (significance level of 5%) are indicated with orange hatching and “*”, and cases where it is less than 0.01 (significance level of 1%) are indicated with yellow hatching and “**”.

For pedestrian accidents, the percentage for all ages was slightly high in single road parts. For students of junior high school age and younger, the percentage of pedestrian accidents on single road parts was higher than that for all ages, with the percentage for students of middle elementary school age and younger being particularly different, with a significance level of 1%. Therefore, pedestrian accidents involving students of middle elementary school age or younger occurred 2–16% more often in single road parts than that involving cases of all ages.

For bicycle accidents, the percentage for all ages was high in intersections. The percentage of bicycle accidents on intersections (excluding that for pre-schoolers) involving students of junior high school age and younger was up to approximately 8% higher than that for all ages, with the percentage for students of middle elementary school age and older being significantly different. Therefore, bicycle accidents involving middle elementary school to junior high school students occurred more often in intersections than those involving all ages.

3.2. Trends of Traffic Accidents Involving Children According to the Number of Lanes

Table 4. Pedestrian accident percentage according to the number of lanes and test results (single-road part).

Age Division	One Lane		Two Lanes		Four Lanes or More		All Accidents (Cases)
	Percentage	p-Value *: p < 0.05 **: p < 0.01	Percentage	p-Value *: p < 0.05 **: p < 0.01	Percentage	p-Value *: p < 0.05 **: p < 0.01
Pre-schooler	33.33%	0.0078 **	65.33%	0.9173	1.33%	0.0001 **	225
Lower elementary school	28.09%	0.3434	70.21%	0.1565	1.70%	0.0001 **	235
Middle elementary school	31.53%	0.1383	66.67%	0.8285	1.80%	0.0082 **	111
Higher elementary school	19.30%	0.3014	75.44%	0.1234	5.26%	0.3234	57
Junior high school	19.40%	0.2716	76.12%	0.0743	4.48%	0.1959	67
All ages	25.29%	1.0000	65.67%	1.0000	9.03%	1.0000	3155

and

Table 5. Pedestrian accident percentage according to the number of lanes and test results (intersections).

Age Division	1 × 1 Lane		2 × 1 Lane		2 × 2 Lane		2 × 4 Lane		4 × 4 Lane		Other	All Accidents (Cases)
	Percentage	p-Value *: p < 0.05 **: p < 0.01	Percentage	p-Value *: p < 0.05 **: p < 0.01	Percentage	p-Value *: p < 0.05 **: p < 0.01	Percentage	p-Value *: p < 0.05 **: p < 0.01	Percentage	p-Value *: p < 0.05 **: p < 0.01	Percentage
Pre-schooler	13.79%	0.2352	28.74%	0.0001 **	40.23%	0.9288	3.45%	0.1450	2.30%	0.0073 **	11.49%	87
Lower elementary school	20.00%	0.0002 **	20.74%	0.0295 *	42.96%	0.4569	3.70%	0.0898	2.22%	0.0008 **	10.37%	135
Middle elementary school	23.44%	0.0004 **	17.19%	0.4721	29.69%	0.1033	4.69%	0.3790	7.81%	0.2692	17.19%	64
Higher elementary school	15.25%	0.1762	20.34%	0.1684	38.98%	0.9047	3.39%	0.2224	11.86%	0.9364	10.17%	59
Junior high school	13.04%	0.3913	14.49%	0.9119	49.28%	0.1106	5.80%	0.5704	7.25%	0.2692	10.14%	69
All ages	9.91%	1.0000	14.02%	1.0000	39.75%	1.0000	7.63%	1.0000	11.53%	1.0000	17.15%	2845

list the pedestrian accident percentages with respect to the number of lanes in single-road parts and intersections, respectively, according to the age group. For all ages, pedestrian accidents more often occurred on two-lane roads for both single-road parts and intersections. For single-road parts, the pedestrian accident percentage in one-lane roads was 3–8% higher for students of middle elementary school age or younger than that for all ages, and a difference with a significance level of 1% was obtained for pre-schoolers.

For intersections, the pedestrian accident percentages at intersections where a one-lane road connects (1 × 1 lane; 2 × 1 lane) were up to approximately 14% higher for students of junior high school age and younger than those of all ages. Particularly, significant differences were obtained for younger elementary school and junior high school students for the 1 × 1 lane as well as for younger elementary school students for the 2 × 1 lane. This may be because the primary movement areas are concentrated on single-lane community roads, such as around residences and on school-commuting roads, where automobile traffic is assumed to be relatively small for walking students of elementary school age and younger.

Table 6. Bicycle accident percentage according to the number of lanes and test results (single-road part).

Age Division	One Lane		Two Lanes		Four Lanes or More		All Accidents (Cases)
	Percentage	p-Value *: p < 0.05 **: p < 0.01	Percentage	p-Value *: p < 0.05 **: p < 0.01	Percentage	p-Value *: p < 0.05 **: p < 0.01
Pre-schooler	41.67%	0.0032 **	55.56%	0.3736	2.78%	0.0316 *	36
Lower elementary school	42.17%	0.0000 **	56.63%	0.2541	1.20%	0.0003 **	83
Middle elementary school	37.35%	0.0005 **	55.42%	0.1726	7.23%	0.0320 *	83
Higher elementary school	26.03%	0.3332	63.01%	0.9665	10.96%	0.2522	73
Junior high school	14.95%	0.0120 *	70.46%	0.0107 *	14.59%	0.5637	281
All ages	21.32%	1.0000	62.77%	1.0000	15.91%	1.0000	2791

and

Table 7. Bicycle accident percentage according to the number of lanes and test results (intersections).

Age Division	1 × 1 Lane		1 × 2 Lane		2 × 1 Lane		2 × 2 Lane		4 × 4 Lane		Other	All Accidents (Cases)
	Percentage	p-Value *: p < 0.05 **: p < 0.01	Percentage	p-Value *: p < 0.05 **: p < 0.01	Percentage	p-Value *: p < 0.05 **: p < 0.01	Percentage	p-Value *: p < 0.05 **: p < 0.01	Percentage	p-Value *: p < 0.05 **: p < 0.01	Percentage
Pre-schooler	43.42%	0.0000 **	2.63%	0.0165 *	19.74%	0.1403	21.05%	0.0460 *	0.00%	0.0129 *	13.16%	76
Lower elementary school	31.40%	0.0000 **	4.13%	0.0004 **	27.69%	0.0000 **	23.97%	0.0104 *	2.07%	0.0014 **	10.74%	242
Middle elementary school	32.94%	0.0000 **	6.27%	0.0112 *	21.57%	0.0005 **	26.27%	0.0645	1.96%	0.0008 **	10.98%	255
Higher elementary school	24.65%	0.0115 *	10.07%	0.4926	21.53%	0.0003 **	28.82%	0.2938	3.47%	0.0099 **	11.46%	288
Junior high school	16.87%	0.2214	11.16%	0.8549	14.21%	0.7855	36.12%	0.0153 *	6.77%	0.4541	14.87%	753
All ages	18.69%	1.0000	11.38%	1.0000	13.85%	1.0000	31.76%	1.0000	7.53%	1.0000	16.79%	6521

list the bicycle accident percentages with respect to the number of lanes in single-road parts and intersections, respectively, according to the age group. For all ages, more bicycle accidents occurred on two-lane roads for both single-road parts and intersections. For single-road parts, the bicycle accident percentage on one-lane roads was 5–21% higher for students of elementary school age or younger than that for all ages, and a difference with a significance level of 1% was obtained for students of middle elementary school age or younger. For intersections, the bicycle accident percentages at intersections where a one-lane road connects (1 × 1 lane, 2 × 1 lane) were up to approximately 25% higher for students of elementary school age and younger than those of all ages. Significant differences were obtained for students of elementary school age and younger for the 1 × 1 lane, as well as for elementary school students for the 2 × 1 lane. This may be because of the same reason explained previously for pedestrians.

3.3. Trends of Traffic Accidents Involving Children with Respect to the Sidewalk/Road Division

Table 8. Pedestrian accident percentage according to the sidewalk/road division and test results (single-road part).

Age Division	Guard Fence		Curb/Block		Roadside Strip		No Division		All Accidents (Cases)
	Percentage	p-Value *: p < 0.05 **: p < 0.01	Percentage	p-Value *: p < 0.05 **: p < 0.01	Percentage	p-Value *: p < 0.05 **: p < 0.01	Percentage	p-Value *: p < 0.05 **: p < 0.01
Pre-schooler	0.44%	0.6233	38.67%	0.0001 **	23.56%	0.0177 *	37.33%	0.0225 *	225
Lower elementary school	0.00%	0.1891	47.23%	0.1686	21.28%	0.1229	31.49%	0.6496	235
Middle elementary school	0.00%	0.3667	45.05%	0.1563	20.72%	0.3512	34.23%	0.3488	111
Higher elementary school	0.00%	0.4831	45.61%	0.3476	15.79%	0.7641	38.60%	0.1652	57
Junior high school	0.00%	0.4551	58.21%	0.3053	10.45%	0.1407	31.34%	0.8234	67
All ages	0.73%	1.0000	51.89%	1.0000	17.31%	1.0000	30.08%	1.0000	3155

and

Table 9. Pedestrian accident percentage according to the sidewalk/road division and test results (intersections).

Age Division	Guard Fence		Curb/Block		Roadside Strip		No Division		All Accidents (Cases)
	Percentage	p-Value *: p < 0.05 **: p < 0.01	Percentage	p-Value *: p < 0.05 **: p < 0.01	Percentage	p-Value *: p < 0.05 **: p < 0.01	Percentage	p-Value *: p < 0.05 **: p < 0.01
Pre-schooler	1.15%	0.8498	60.92%	0.0001 **	11.49%	0.4692	26.44%	0.0000 **	87
Lower elementary school	0.00%	0.2555	63.70%	0.0001 **	13.33%	0.1086	22.96%	0.0001 **	135
Middle elementary school	0.00%	0.4336	70.31%	0.1276	12.50%	0.3696	17.19%	0.1661	64
Higher elementary school	0.00%	0.4522	66.10%	0.0254 *	8.47%	0.8467	25.42%	0.0011 **	59
Junior high school	0.00%	0.4162	73.91%	0.3858	8.70%	0.8840	17.39%	0.1368	69
All ages	0.95%	1.0000	78.28%	1.0000	9.21%	1.0000	11.56%	1.0000	2845

list the pedestrian accident percentages with respect to the sidewalk/road division in single-road parts and intersections, respectively, according to the age group. For all ages, pedestrian accidents occurred more often on roads with curbs/blocks as divisions for both single-road parts and intersections. For single-road parts, the pedestrian accident percentage on roads without roadside strips or sidewalks was up to approximately 7% higher for students of middle elementary school age or younger than that for all ages; however, a significant difference was obtained only for pre-schoolers. Therefore, pedestrian accident trends in single-road parts with respect to the sidewalk/road classification might be approximately the same between children and all-age groups. For intersections, the pedestrian accident percentages in roads with no sidewalk/road division were 6–15% higher for students of junior high school age and younger than those for all ages, and a difference with a significance level of 1% was obtained for students of elementary school age and younger (excluding middle elementary school students). This may be because the primary movement areas of walking for students of elementary school age and younger are concentrated on community roads without sidewalk/road divisions.

Table 10. Bicycle accident percentage according to the sidewalk/road division and test results (single-road part).

Age Division	Guard Fence		Curb/Block		Roadside Strip		No Division		All Accidents (Cases)
	Percentage	p-Value *: p < 0.05 **: p < 0.01	Percentage	p-Value *: p < 0.05 **: p < 0.01	Percentage	p-Value *: p < 0.05 **: p < 0.01	Percentage	p-Value *: p < 0.05 **: p < 0.01
Pre-schooler	0.00%	0.5532	25.00%	0.0000 **	30.56%	0.0004 **	44.44%	0.0003 **	36
Lower elementary school	0.00%	0.3680	33.73%	0.0000 **	13.25%	0.6151	53.01%	0.0000 **	83
Middle elementary school	1.20%	0.8282	44.58%	0.0000 **	15.66%	0.2391	38.55%	0.0000 **	83
Higher elementary school	1.37%	0.7301	52.05%	0.0055 **	16.44%	0.1901	30.14%	0.0347 *	73
Junior high school	0.71%	0.6727	72.60%	0.0810	10.68%	0.6914	16.01%	0.1037	281
All ages	0.97%	1.0000	67.50%	1.0000	11.47%	1.0000	20.06%	1.0000	2791

and

Table 11. Bicycle accident percentage according to the sidewalk/road division and test results (intersections).

Age Division	Guard Fence		Curb/Block		Roadside Strip		No Division		All Accidents (Cases)
	Percentage	p-Value *: p < 0.05 **: p < 0.01	Percentage	p-Value *: p < 0.05 **: p < 0.01	Percentage	p-Value *: p < 0.05 **: p < 0.01	Percentage	p-Value *: p < 0.05 **: p < 0.01
Pre-schooler	0.00%	0.4674	26.32%	0.0000 **	14.47%	0.4913	59.21%	0.0000 **	76
Lower elementary school	0.00%	0.1948	32.64%	0.0000 **	16.12%	0.0479 *	51.24%	0.0000 **	242
Middle elementary school	0.39%	0.5698	29.80%	0.0000 **	16.08%	0.0444 *	53.73%	0.0000 **	255
Higher elementary school	0.69%	0.9930	44.79%	0.0000 **	15.97%	0.0379 *	38.54%	0.0000 **	288
Junior high school	1.06%	0.2553	57.64%	0.0318 *	12.35%	0.7181	28.95%	0.0580	753
All ages	0.69%	1.0000	61.66%	1.0000	11.90%	1.0000	25.75%	1.0000	6521

list the bicycle accident percentages with respect to the sidewalk/road division in single-road parts and inter-sections, respectively, according to the age group. For all ages, similar to the trend observed for pedestrian accidents, bicycle accidents occurred more often on roads with curbs/blocks as divisions for both single-road parts and intersections. For children, in both single-road parts and intersections, the bicycle accident percentages on roads with no sidewalk/road division were up to approximately 34% higher for students of elementary school age and younger than those of all ages. In addition, a significant difference was obtained for students of elementary school age and younger. This may be because of the same reason explained previously for pedestrians.

3.4. Trends of Traffic Accidents Involving Children According to the Speed Limit

Table 12. Pedestrian accident percentage according to the speed limit and test results (single-road part).

Age Division	30 km/h or Less		40 km/h		50 km/h		No Speed Limit		Other	All Accidents (Cases)
	Percentage	p-Value *: p < 0.05 **: p < 0.01	Percentage	p-Value *: p < 0.05 **: p < 0.01	Percentage	p-Value *: p < 0.05 **: p < 0.01	Percentage	p-Value *: p < 0.05 **: p < 0.01	Percentage
Pre-schooler	20.00%	0.0005 **	31.11%	0.0400 *	6.22%	0.0007 **	41.78%	0.0352 *	0.89%	225
Lower elementary school	20.85%	0.0001 **	39.57%	0.6254	6.38%	0.0007 **	32.77%	0.5206	0.43%	235
Middle elementary school	16.22%	0.1905	38.74%	0.8700	5.41%	0.0080 **	38.74%	0.3964	0.90%	111
Higher elementary school	19.30%	0.0988	42.11%	0.5241	3.51%	0.0204 *	35.09%	0.9682	0.00%	57
Junior high school	14.93%	0.4797	40.30%	0.6978	14.93%	0.8840	29.85%	0.3966	0.00%	67
All ages	12.08%	1.0000	37.97%	1.0000	14.29%	1.0000	34.83%	1.0000	0.82%	3155

and

Table 13. Pedestrian accident percentage according to the speed limit and test results (intersections).

Age Division	30 km/h or Less		40 km/h		50 km/h		No Speed Limit		Other	All Accidents (Cases)
	Percentage	p-Value *: p < 0.05 **: p < 0.01	Percentage	p-Value *: p < 0.05 **: p < 0.01	Percentage	p-Value *: p < 0.05 **: p < 0.01	Percentage	p-Value *: p < 0.05 **: p < 0.01	Percentage
Pre-schooler	17.24%	0.0024 **	36.78%	0.0841	12.64%	0.2539	33.33%	0.2739	0.00%	87
Lower elementary school	17.78%	0.0001 **	45.19%	0.8259	12.59%	0.1534	24.44%	0.3705	0.00%	135
Middle elementary school	17.19%	0.0090 **	45.31%	0.8941	15.63%	0.7215	21.88%	0.2813	0.00%	64
Higher elementary school	8.47%	0.9134	32.20%	0.0333 *	20.34%	0.5460	38.98%	0.0630	0.00%	59
Junior high school	4.35%	0.2582	53.62%	0.2187	17.39%	0.9892	23.19%	0.3805	1.45%	69
All ages	8.08%	1.0000	46.15%	1.0000	17.33%	1.0000	27.98%	1.0000	0.46%	2845

list the pedestrian accident percentages with respect to the speed limit in single-road parts and intersections, respectively, according to the age group. For all ages, pedestrian accidents occurred more often on roads with a speed limit of 40 km/h for both single-road parts and intersections. For single-road parts, the pedestrian accident percentage on roads with speed limits of 30 km/h or lower was 3–9% higher for students of junior high school age or younger than that for all ages, and a difference with a significance level of 1% was obtained for students of lower elementary school age or younger. For intersections, the pedestrian accident percentages on roads with a speed limit of 30 km/h or lower were higher for students of elementary school age or younger than those for all ages, and a difference with a significance level of 1% was obtained for students of middle elementary school age or younger. This may be because the primary movement areas of walking for students of elementary school age and younger are concentrated on community roads with low-speed limits or narrow town streets.

Table 14. Bicycle accident percentage according to the speed limit and test results (single-road part).

Age Division	30 km/h or Less		40 km/h		50 km/h		No Speed Limit		Other	All Accidents (Cases)
	Percentage	p-Value *: p < 0.05 **: p < 0.01	Percentage	p-Value *: p < 0.05 **: p < 0.01	Percentage	p-Value *: p < 0.05 **: p < 0.01	Percentage	p-Value *: p < 0.05 **: p < 0.01	Percentage
Pre-schooler	16.67%	0.0293 *	25.00%	0.2864	5.56%	0.0760	52.78%	0.2069	0.00%	36
Lower elementary school	13.25%	0.0361 *	30.12%	0.5286	2.41%	0.0005 **	54.22%	0.0307 *	0.00%	83
Middle elementary school	14.46%	0.0123 *	34.94%	0.7738	7.23%	0.0230 *	42.17%	0.9788	1.20%	83
Higher elementary school	16.44%	0.0028 **	32.88%	0.9213	16.44%	0.9727	34.25%	0.1681	0.00%	73
Junior high school	6.76%	0.8018	38.08%	0.1164	16.37%	0.9250	38.79%	0.2539	0.00%	281
All ages	7.17%	1.0000	33.43%	1.0000	16.59%	1.0000	42.31%	1.0000	0.50%	2791

and

Table 15. Bicycle accident percentage according to the speed limit and test results (intersections).

Age Division	30 km/h or Less		40 km/h		50 km/h		No Speed Limit		Other	All Accidents (Cases)
	Percentage	p-Value *: p < 0.05 **: p < 0.01	Percentage	p-Value *: p < 0.05 **: p < 0.01	Percentage	p-Value *: p < 0.05 **: p < 0.01	Percentage	p-Value *: p < 0.05 **: p < 0.01	Percentage
Pre-schooler	13.16%	0.5250	31.58%	0.4484	7.89%	0.3540	46.05%	0.5914	1.32%	76
Lower elementary school	19.01%	0.0001 **	24.79%	0.3263	5.79%	0.0076 **	49.59%	0.8936	0.83%	242
Middle elementary school	22.75%	0.0000 **	21.18%	0.0227 *	5.10%	0.0020 **	49.80%	0.8374	1.18%	255
Higher elementary school	15.97%	0.0070 **	28.82%	0.6682	5.56%	0.0024 **	48.96%	0.9495	0.69%	288
Junior high school	10.36%	0.6672	27.09%	0.7392	10.23%	0.3884	51.26%	0.2722	1.06%	753
All ages	10.87%	1.0000	27.66%	1.0000	11.27%	1.0000	49.15%	1.0000	1.04%	6521

list the bicycle accident percentages with respect to the speed limit in single-road parts and intersections, respectively, according to the age group. For all ages, bicycle accidents occurred more often on roads without speed limits for both single-road parts and intersections. For single-road parts, the bicycle accident percentage on roads with a speed limit of 30 km/h or less was 6–9% higher for students of elementary school age or younger than that for all ages, and a significant difference was obtained. For intersections, trends similar to those observed for single-road parts were obtained, and a difference with a significance level of 1% was obtained for elementary school students. This may be due to the same reason explained previously for pedestrians.

3.5. Road Structures in Which Traffic Accidents That Involve Children Occur

Table 16. Road structures in which pedestrian accidents that involve children occur. *: p < 0.05 **: p < 0.01.

Age Division	Road Shape	Number of Lanes		Sidewalk/Road Division		Speed Limit
		Single Road Part	Intersection	Single Road Part	Intersection	Single Road Part	Intersection
Pre-schooler	Single road part **	1 lane **	2 × 1 lane **	Roadside strip *	No division **	30 km/h or less **	30 km/h or less **
Lower elementary school	Single road part **	2 lanes	1 × 1 lane **	Roadside strip	No division **	30 km/h or less **	30 km/h or less **
Middle elementary school	Single road part **	1 lane	1 × 1 lane **	No division	No division	30 km/h or less	30 km/h or less **
Higher elementary school	Intersection *	2 lanes	2 × 1 lane	No division	No division **	30 km/h or less	No speed limit
Junior high school	Intersection	2 lanes	2 × 2 lane	Curb/block	No division	No speed limit	40 km/h

and

Table 17. Road structures in which bicycle accidents that involve children occur. *: p < 0.05 **: p < 0.01.

Age Division	Road Shape	Number of Lanes		Sidewalk/Road Division		Speed Limit
		Single Road Part	Intersection	Single Road Part	Intersection	Single Road Part	Intersection
Pre-schooler	Intersection	1 lane **	1 × 1 lane **	No division **	No division **	30 km/h or less *	40 km/h
Lower elementary school	Intersection	1 lane **	2 × 1 lane **	No division **	No division **	No speed limit *	30 km/h or less **
Middle elementary school	Intersection *	1 lane **	1 × 1 lane **	No division **	No division **	30 km/h or less *	30 km/h or less **
Higher elementary school	Intersection **	1 lane	2 × 1 lane **	No division *	No division **	30 km/h or less **	30 km/h or less **
Junior high school	Intersection *	2 lanes *	2 × 2 lane	Curb/block	No division	40 km/h	No speed limit

list the road structures in which traffic accidents that involve children occur in pedestrians and bicycles, respectively, according to the age group. For pedestrian accidents there are many pedestrian accidents involving students of middle elementary school age or younger in single road parts for the road shape, intersections where a one-lane road connects for the number of lanes, intersections with speed limits of 30 km/h or less, and a difference with a significance level of 1% was obtained. In addition, there are many pedestrian accidents involving students of elementary school age and younger (excluding middle elementary school students) on roads with no sidewalk/road division, and a difference with a significance level of 1% was obtained. Therefore, road structures in which pedestrian accidents that involve students of middle elementary school age or younger occur consist of a one-lane road with a speed limit of 30 km/h or lower and no sidewalk/road division.

For bicycle accidents there are many bicycle accidents involving middle elementary school to junior high school students in intersections for the road shape, and a significant difference was obtained. There are many bicycle accidents involving elementary school age or younger students on one-lane roads for the number of lanes, and a difference with a significance level of 1% was obtained for students of elementary school age and younger (excluding higher elementary school students in single road parts). There are many bicycle accidents involving students of elementary school age and younger (excluding both lower elementary school students in single road parts and pre-schoolers in intersections) on roads with speed limits of 30 km/h or lower for the speed limit, and a significant difference was obtained. In addition, there are many bicycle accidents involving students of elementary school age and younger on roads with no sidewalk/road division, and a significant difference was obtained. Therefore, road structures in which bicycle accidents involve students of elementary school age and younger occur. This consists of a one-lane road with a speed limit of 30 km/h or lower and no sidewalk/road division.

4. Conclusions

This study aimed to clarify the road structure in which traffic accidents that involve walking or bicycle-riding children occur. In this regard, traffic accident trends based on the age of children and road structure were analyzed using statistical data on traffic accidents obtained from the Ishikawa Prefectural Police Headquarters. In performing of the analysis, we used the type and age of each party involved and the road structure (road shape, number of lanes, sidewalk/road division, and speed limit). The number and percentage of pedestrian/cyclist-related accidents were calculated for age groups, and then the accident percentage of each age group was compared with that of all ages. We also judged significant differences using the Hypothesis Testing for the Difference in the Population Proportions.

Our results enabled the determination of the road structure in which pedestrian and bicycle accidents involving children were more likely to occur. This consists of a one-lane road with a speed limit of 30 km/h or lower and no sidewalk/road division. These road structures are assumed to be common on community roads with low standards and priority for pedestrians, such as around residences and on school-commuting roads. Therefore, to reduce pedestrian and bicycle accidents involving students of elementary school age or younger, it is important to raise traffic safety awareness among automobile drivers who use community roads. In addition, we think that using humps, smooth pedestrian crossings, narrow fences, chicane, and the slalom are effective as the maintenance of pedestrians and the sidewalk and the protective fence separating the automobile physically, the speed restraint measures of the automobile.

Road structures with many traffic accidents of children in Ishikawa were clarified in this study. However, there is a problem in that there are few samples of the pedestrian accidents in particular because we subdivided age division. In addition, because the results in this study were obtained from statistical data on traffic accidents obtained from the Ishikawa Prefecture, other prefectures will have different elementary-school-student commuting patterns as well as community road and town street road structures. Therefore, traffic accident trends might differ. In the future, the scope of analysis should include other regions and the entire country. Additionally, it is necessary to carry out a robustness test when we make a model. Furthermore, although road structural parameters, such as the road shape, number of lanes, and sidewalk/road division, can be generally obtained from the data, these do not consider traffic conditions, such as the traffic volume of automobiles, bicycles, and pedestrians or the average travel speed on roads. Therefore, future studies should consider these factors.