The development of transportation systems in Thailand has been carried out to support the expansion of cities and facilitate domestic transportation. Due to the first-car policy of Yingluck Shinawatra’s government, the number of private cars increased by 169,861 during 2011-2012 [1]. Although it helped stimulated the state economy, the increase of private cars and motorcycles was found to cause traffic problems, energy wastage, and increased levels of accidents [2]. Traffic accidents are one of Thailand’s main problems [3]. Student drop-off and pick-up at schools is a key cause of traffic congestion in urban areas of Thailand [4]. Thailand is a country where primary schools have varying degrees of quality and reputation [5]. In addition, drop-off and pick-up behaviors of Thai parents are different from those of parents in other countries. Thai parents tend to be concerned about traffic problems and school environment that can affect the safety of their children. These factors directly affect children safety at a national level. Apart from physical factors and drop-off and pick-up behaviors mentioned earlier, recent studies also suggested that parents’ perception of hazardous environments can affect their routine behavior and opinions such as driving into school or in front of classrooms, which may cause harm to students [6, 7, 4]. This research aimed to examine the risk factors associated with school drop-off and pick-up in three dimensions: 1) risk zones, 2) risk behaviors, and 3) risk periods. The data obtained from 5 primary schools in Thailand, which were selected as the case studies, were compared and analyzed. The spatial analysis of traffic accidents was conducted using the actual accident statistics. Moreover, a questionnaire survey was carried out on the parents and teachers from Thai kindergartens and primary schools in Thailand in order to obtain data about drop-off and pick-up behaviors and opinions on children’s risk factors.

Childhood is an important period that basically requires safety [8]. As children aged 4-12 years [9] have no abilities to care for their own safety and protect themselves from danger caused by unfamiliar environment, their parents or teachers have a responsibility to take care of them. Children will develop a sense of safety and perception of risk through learning from parents, family, and environment during the early period of their life [10-12]. This demonstrates that children’s positive or negative perception towards self-protection from danger depends mainly on their childhood experience and environment [13, 14]. This is consistent with a Californian study about environmental improvements for walking, sidewalks and traffic control, which suggested that children walking to school are significantly more likely to feel confident when walking and cycling facilities are safe and improved [15].

This study aimed to investigate the risk zones around schools by focusing on the accident-prone areas outside schools and inside-school environment based on the parents’ opinions obtained from the questionnaire.

Although Thailand has rules and regulations about land use, transportation systems, and public utility systems, there is still a lack of effective urban management and spatial planning [19] associated with public services such as schools and hospitals. There is no catchment area within residential zones [20]. Furthermore, the quality and reputation levels of primary schools in Thailand are enormously varied. Thai parents generally want to have their children study at the primary schools with higher quality. Anuban School is a provincial primary school with high academic standards and reputation. It is located in the downtown area and usually causes traffic problems in nearby areas. The parents, who are concerned about the safety of the school environment, are likely to drop-off and pick-up their children at school by themselves [21]. Thus, it can be said that the parents’ drop-off and pick-up behaviors have a direct impact on the safety of children [22-24]. This study focused to explore the drop-off and pick-up behaviors, methods, and areas, including the types of vehicles used by the parents.

Time and situation have an effect on the drop-off and pick-up behaviors. The parents’ drop-off and pick-up behaviors and methods in the morning and after school are different. The differences in perception of danger in the morning and after school affect the way the parents and teachers take care of the children [5]. In this study, an in-depth investigation was carried out. The questionnaire was used to elicit the opinions of the parents and teachers on the risk zones and risk periods. The results would be compared and analyzed in order to identify the high-risk periods for children.

Based on the literature review and past studies, the scope and framework of this study can be summarized into 4 parts: 1) traffic accidents, 2) school environment, 3) children’s travel behaviors, and 4) opinions of parents and teachers.

Fig. (1). The relationship between variables in the study: adapted from [6, 7, 25] designed by the authors.

Traffic accidents are empirically dangerous incidents that often occur with children around school areas. This is probably because the children spend most of their time at school, their school is located in traffic congestion area, most parents arrive and depart the school at the same time during rush hours, and the parents’ perception about the danger of school environment. Moreover, there are four important variables that affect travel behavior, which are 1) socio-economic condition: previous studies suggested that middle- to high-income families are more aware of and look out for the safety of children than low-income families, 2) individual opinions: recent studies indicated that anxiety about danger has an effect on an individual’ behavior patterns, 3) lifestyle: past research revealed that the burden of family affects children’s travel method, for example, the parents with burdens tend to let their children take public buses to school or have the relatives, especially grandmothers, send their children to school instead of themselves [4], and 4) school environment [6, 7, 24]. In addition, as parents and teachers are the main caretakers of children, understanding their attitudes and opinions will be helpful in identifying related risk factors. The details are shown in Fig. (1).

Thailand is divided into 77 provinces in 6 regions: 1) northern region (9 provinces), 2) central region (21 provinces), 3) northeastern region (20 provinces), 4) eastern region (7 provinces), 5) western region (5 provinces), and 6) southern region (14 provinces). Early childhood education in Thailand is under the supervision of the Ministry of Education. Anuban Schools are high-quality public schools that accept both kindergarten and primary school students. They are managed as an association in the form of Provincial Anuban Association. There is only one Anuban School in each province. In order to make sure that the results could be generalized to schools across the country, the researcher selected to carry out the present study in Anuban Schools because of the following four reasons. Firstly, they are located in downtown with urban contexts. Secondly, they have a similar level of quality and reputation. Thirdly, the parents of students in Anuban Schools have the similar socio-economic background and family lifestyle. Fourthly, they offer both kindergarten and primary school education in the same school [4, 5]. The researcher selected the 5 case study schools from 5 regions, excluding the central region where Bangkok is a center because it has a completely different context from other regions. The following two factors were taken into account during the case study selection process: 1) the statistics of accidents occurring to children in each province, the statistics of accidents occurring to children within a 500-meter radius of schools, and the statistics of accidents occurring during the drop-off and pick-up time were collected from 65 Anuban Schools in order to find out the schools with the highest accident record in each 5 region [5], and 2) the physical environment of schools that could cause harm to children. The 5 Anuban Schools that were selected as the case studies consisted of Anuban Lampang School from the northern region, Anuban Nakhon Ratchasima School from the northeastern region, Anuban Phuket School from the southern region, Anuban Chonburi School from the eastern region, and Anuban Kanchanaburi School from the western region.

Kernel Density Estimation, which is a nonparametric spatial analysis using Geographic Coordinate Systems (GIS), was used to find out the frequency of coordinates [26-28] that could identify the intensity of accident factors or spatial opinions. This technique is highly popular in urban planning, transportation planning and spatial studies [29-34]. The spatial data analyzed by ArcGIS will be shown in form of raster. The principle of this method is to calculate the radius of each point and link that point to another using bandwidth. The radius of the circular neighborhood has an effect on the resulting density map. If the radius is increased, there is a possibility that the circular neighborhood may include more feature points, which results in a smoother density surface [35, 36]. The cell size chosen was 1 m × 1 m.

The Road Safety Culture is a private organization that deals with road accident reports for 24 hours and keeps a detailed report that shows the statistics of accidents occurring at different time throughout the country. This data is highly accurate since it is referred from the database of insurance companies and car accident insurance firms [37]. Every registered car is required to have insurance. The statistical data of accidents have been continually collected for years. The most important issue is that the accident coordinates are identified in form of GIS geographic maps [26, 27], which were presented only at the national level. In previous research studies, most statistical data analyzed included the statistics of accidents occurring to children, accidents occurring to children within a 500-meter radius around schools, and accidents occurring to children during the drop-off and pick-up time. However, the GIS data of accident locations have not yet been spatially analyzed to identify the risk factors that can harm children [5]. Thus, Kernel Density Analysis was applied to analyze the collected data in the present research.

In addition, the data was divided into 2 parts, consisting of the coordinates of accidents occurring around the school zone within a 500-meter radius and the coordinates of risk zones in the parents’ opinions. The parents were asked to identify the highest risk zones and risk periods using the questionnaire. The data obtained from 1,230 questionnaires were compared and analyzed to find out actual accident-prone locations based on the parents’ opinions. The resulting coordinates were shown in the Shape File format (.shp). The data were divided into 10 equal layers. Bilinear Interpolation was applied to demonstrate the results. The image quality was set to normal with 30% transparency. The criteria for identifying the children’s risk zones were 1) the overall condition of the accident locations, including both the highest and lowest accident areas. This was because an accident could cause more severe and dangerous impacts on children than adults, due to the children’s physical weakness; 2) the dangerous areas based on the parents’ opinions, and 3) the overlapping areas between 1 and 2. The spatial results would be used to monitor the safety of children through the safety designing, management, and prevention. The quadratic Kernel function was applied in the present research as follows [35].

(1)

where:

f(x, y): the density estimates at the accident location (x, y),

n: the number of observations,

h: the bandwidth or kernel size,

k: the kernel function, and

d_i: the distance between the accident’s location (x, y) and the location of the i observation.

The descriptive statistical analysis was used to analyze the data obtained from the questionnaires such as frequency, mean, percentage, standard deviation. The inferential statistical analysis was used to analyze the opinions of the parents and the teachers in order to find the differences between the results with t-test and Anova. The empirical data obtained from the physical environment survey and the interview were used to examine and discuss the results of the spatial analysis and the statistical analysis of the data obtained from the questionnaires as shown in Fig. (2).

Fig. (2). The research methodology.

The case study schools could be divided into 2 groups according to the location: 1) Anuban Schools located in the downtown areas with traffic congestion during the drop-off and pick-up time, and 2) Anuban Schools located in the government office areas with advantages of facilities such as Anuban Chonburi School that is located next to a public park and a city hall and Anuban Phuket School that is located next to a temple, where the parents could park their cars in during the drop-off and pick-up time.

In addition, the case studies could be divided into the following 3 groups based on the types of roads surrounding their location: 1) Anuban Schools located on the high way, where the traffic was heavy and full of high-speed vehicles; 2) Anuban Schools located on the main road; and 3) Anuban Schools located on the local road, where the traffic is light and difficult to access such as Anuban Kanchanaburi School.

Considering the frequency of accidents occurring around schools and the spatial results of the Kernel Density Analysis regarding, it was found that the accidents took place in 2 forms, which were 1) cluster: the accidents occurred in cluster, and 2) linear: the accidents occurred in longitudinal distribution. When analyzing the highest accident areas together with the school environment and the behaviors, the results could be summarized as follows: 1) the accidents usually occurred at the junction or a roundabout near the schools, and 2) speed, traffic density, sizes of roads have an impact on the occurrence of accidents around the drop-off and pick-up areas (Fig. 3 and Table 1).

Fig. (3). Accident crash zone within school zone.

Fig. (4). The parents’ opinions toward risk zone.

Regarding the risk zones for children, the results obtained from 1,230 questionnaires revealed that the parents thought that the most dangerous areas for children were the crossing points, the drop-off and pick-up points at the school entrance, and the sidewalk in front of the schools. The risk areas tended to cluster together (Table 2). The drop-off and pick-up of kindergarten students and primary school students were carried out at different time and places, depending on the policy of each school. The number of dangerous areas in each school seemed to correlate with the number of drop-off and pick-up points. The drop-off and pick-up pattern of all case study schools were similar except Anuban Kanchanaburi School. With only one main entrance, Anuban Kanchanaburi School allowed the students to freely go out of the school and wait for their parents outside because of 2 reasons: 1) the school was located in a narrow alley that was difficult for cars to access, and 2) the students needed to buy snack and food outside the school. Based on the parents’ opinions, there were two risk zones for children, which included the area in front of the school on the local road and the main drop-off and pick-up point nearby the junction. It could not be ensured that the students of Anuban Kanchanaburi School would be safer than the students in other schools (Fig. 4).

When comparing the actual accident locations around the schools (Fig. 3) with the risk zones for children in the parents’ opinions (Fig. 4), it was found that the accident locations overlapped with the risk zones, especially in the areas around the drop-off and pick-up points. The accidents often occurred when the students crossed the road. This indicated that the traffic signs warning vehicle users to be more careful and use appropriate speed were ineffective (Figs. 5-9).

Fig. (5). Overlap of external and internal risk zones around Anuban Lampang.

Fig. (6). Overlap of external and internal risk zones around Anuban Nakhon Ratchasima.

Fig. (7). Overlap of external and internal risk zones around Anuban Phuket.

Fig. (8). Overlap of external and internal risk zones around Anuban Chonburi.

Fig. (9). Overlap of external and internal risk zones around Anuban Kanchanaburi.

Regarding the parents’ opinions about the risk zones associated with 8 components of the school environment, the parents stated that the sidewalks and the crossing points were the highest risk zone (49.80 percent), which was in line with the spatial analysis results (Figs. 5-10). The drop-off and pick-up points, school gates, and school fences were the areas with the second highest risk (18.08 percent). Moreover, 9.89 percent of the parents stated that the circulation route within the schools, including the roads, pavements, and stairs, was also one of the risk zones. The opinions of the parents from each individual school were consistent (Fig. 11). Only the parents from Anuban Kanchanaburi School stated that the sidewalks and the crossing points were the highest risk zones (63.22 percent) because the drop-off and pick-up points were outside the school, which was very dangerous for the students both in the morning and afternoon.

Fig. (10). Families characteristics and the drop-off and pick-up behaviors of the parents.

Fig. (11). A comparison between the parents’ and teachers’ opinions about the risk zones.

Considering the teachers’ opinions about highest risk zones, it was found that the teachers’ opinions were consistent with the parents’ opinions (Fig. 12). However, the parents’ opinions were different from the teachers’ opinions in 2 aspect as follows. First, some parents were worried about the blind spots behind the school buildings (4.83 percent), whereas the teachers thought that those blind spots were not dangerous (1.28 percent) because the teachers were more familiar with those areas. When separately comparing the opinions of the parents and the teachers in each individual school, it was found that 3.90 percent of the teachers at Anuban Chonburi School stated that the blind spots behind the school buildings were the risk zones while only 0.33 percent of the parents thought that those areas were dangerous. This was probably because there were a lot of blind spots in the school. Second, in the afternoon the teachers allowed the students to play and wait for the parent freely, which could be the cause of accidents and child missing. Due to a large number of children, it might difficult to find and identify the missing child. Therefore, the teachers stated that the open-playgrounds were one of the risk zones (7.37 percent), whereas the parents did not thought that such areas were the risk zones (1.43 percent).

Fig. (12). A comparison between the parents’ and teachers’ opinions about the risk periods.

In terms of the parents’ opinions about the risk periods for children, it was found that the highest risk period for children was during the drop-off and pick-up time (42.97 percent). The second highest risk period for children was the after-school time or during 04.00 p.m. - 06.00 p.m. (39.87 percent). It was found that the opinions of the parents from each 5 schools were consistent. However, the parents of Anuban Nakhon Ratchasima School stated that the after-school time was the highest risk period for children (53.14 percent). This was because Anuban Nakhon Ratchasima School was the largest school among the case studies. There was a main road with heavy traffic in front of the school and a limited space for the drop-off and pick-up activity.

Considering the teachers’ opinions about the risk periods for children, it was found that most of the teachers stated that the highest risk period for children was during the drop-off and pick-up time (42.63 percent), which was consistent with the parents’ opinions. The teachers also thought that the before-class time (06.00 a.m. - 08.00 a.m.) was the second highest risk period for children (36.54 percent). Moreover, it was found that the teachers from all schools had consistent opinions, except the teachers from Anuban Nakhon Ratchasima School. They stated that the before-class time was the highest risk period for children (48.44 percent). Some teachers from Anuban Kanchanaburi School stated that the after-school time was the risk period for children due to the children’s behaviors that wanted to buy snacks and wait for their parents at the meeting points outside the school.

A comparison between the parents’ and the teachers’ opinions about the risk zones and the risk period could be summarized as follows. It could be seen that the parents’ and the teachers’ opinions about the risk zones for children were consistent. They stated that the sidewalks and the crossing points were the highest risk zones for children. As for the parents’ and the teachers’ opinions about the risk periods for children, it was found that their opinions were significantly different at p < .01 (Table 3).

The parents stated that the after-school time was the risk period for children because it was in the evening and the pick-up points were scattered across different areas both inside and outside the school. It was the parents’ responsibility to look out for the safety of children when returning home from school. The teachers clearly believed that the before-class time was the risk period for children because it was during the rush hours and the traffic was heavier than in the evening. Furthermore, the teachers were the main persons responsible for the children during school time.