Perceptions of Parents of the Quality of the Public Transport Services Used by Children to Commute to School

: The risk of accidents is a danger in public transport that could lead to threats to property, the environment, and the lives and health of people. In particular, parents are keenly aware of and concerned about the safety of school trips. Thus, this study aims to examine the factors that inﬂuence the perceptions of parents about the safety of the school trips of children. The study recruited 750 respondents from Northeast Thailand. Data were obtained from responses to a self-report questionnaire. The model consisted of six factors, namely, transportation satisfaction, infrastructure, information, the safe behavior of drivers, the safety of the transportation systems, and the safety policy. The results of the conﬁrmatory factor analysis indicated that all six factors were in accordance with the empirical data (model ﬁt statistic: χ 2 = 758.098, df = 276, χ 2 /df = 2.747, CFI = 0.962, TLI = 0.955, SRMR = 0.038, RMSEA = 0.048). The results can serve as a reference for developing guidelines and recommending policies for the management and the implementation of safe school trips for students.


Introduction
One of society's major concerns in Thailand is the school commute, a critical topic for households with students. In many countries worldwide, students frequently drive their own cars to school, for example, see Queiroz, et al. [1]; this increase in the overall number of cars results in school areas being congested with vehicles, a common scenario in Thailand's urban areas. This usually occurs at peak times at the start and end of a school day. School buses and cars are the most common modes of transport for children to and from school [2]. Even though students are at risk of accidents, the Road Safety Thailand Road Safety Policy Foundation [3] reported that students are approved to use a private vehicle to travel to school because of the ease of travel, reduced time pressure [4][5][6][7], and the perceived safety of parents in traveling to and from school [8][9][10][11]. Parents are worried about the safety of children traveling to school due to the increasing number of traffic incidents. In 2020, the death rate due to road accidents was 2720 per 100,000 people, with the highest rate among children and young adults aged 10 to 24 years [12]. The Public Health Statistics [13] reported that Thai children aged less than 15 years die from injuries sustained on the roads at a rate of 93.7 deaths per 100,000 people.
This situation demonstrates the importance of public transportation as an essential component of urban infrastructure that helps people move around and supports the economy [14]. Previous studies, such as those conducted by Dianat, et al. [15], Quiñonez,

Conceptualization of the Effect of Parenting Styles on Children's Risk Perceptions on Roads
There is evidence that parenting styles affect children's perceptions of the risk of travel. In other words, children tend to imitate the behavior of their parents by following their choices and attitudes [1]. It is therefore expected that if parents are aware of the impact of the perceived risks on the road, this can improve their children's perceptions of risk. However, the results have indicated that parents' perceptions of traffic risk were not significantly related to children's perceptions of risk [39]. Parents' perceived safety Sustainability 2022, 14, 13005 3 of 20 in transport operations and services affects parents' trust in their children's solo travel on public transportation as well as the latent concerns related to "operation service". For example, frequency and schedule greatly affect the willingness to switch to public transport on school trips [1]. In addition, parents' perceptions of safety in their children's travel to school are influenced by age, income, and social cohesion, as well as the quality of the infrastructure [40].

Measurement of the Perception of Transport Users of the Quality of Public Transportation
According to previous studies, Morton, et al. [41] used exploratory factor analysis (EFA) to classify relevant indicators and confirmatory factor analysis (CFA) to verify the consistency of the grouping of indicators with the empirical data in a study on the perception of transport users of the quality of public transportation. The related factors included comfort, convenience, punctuality, and access to information. The current study investigated the factors that influence the perceptions of parents about the safety of the school trips of children. The factors expected to improve safety and convenience in travel can also be used to formulate guidelines for the safety management of school trips for students and to recommend policies for the safety of school travel for students. Therefore, six factors were applied as follows:

Transportation
The awareness of parents about the safety of school trips is a factor related to the indicators in terms of selecting the mode of public transport for school trips, such as access to school transportation services, school commute facilities, and planned schedule management. The majority of studies that focus on the factors of service availability cite that the availability of routes and clearly stated schedules of services exerted a significant influence on the perceptions of the quality of the public transport services [41][42][43][44][45][46][47]. Moreover, travel time exerts a negative influence on user satisfaction [48].

Infrastructure
The service infrastructure of the public transport for schools is an extremely important factor for education, because consumers mainly pay attention to infrastructure when selecting a mode of transport. In addition, it exerts a positive influence on the perceived quality of the public transport services [41,44,45,47]. A factor related to the indicators of infrastructure is comfort at bus stops, such as having a roof to protect commuters from rain, sufficient lighting, cleanliness, and sufficient seats for waiting. Previous studies confirm that the convenience of access to public bus stops exerted a statistically significant effect on the satisfaction of users of public transport.

Information
Public bus operators should prioritize travel information services to increase the positive perceptions of consumers about the service quality [46]. Public transportation systems that introduce technology systems to facilitate travel information services can respond immediately to travel planning, including the establishment of an information center to cater to questions and for public relations. Regarding public transport to schools, numerous studies find that a service that is comprehensive, accurate, up-to-date, and capable of meeting the needs of users is a factor that influences the perception of the quality of the service. Moreover, the studies cite that the overall service of public transport was statistically significant [41][42][43][44]49].

Safety Behavior of Drivers
On the road, drivers should avoid behaviors that lead to risk and harm for themselves and others. For example, van Lierop and El-Geneidy [49] identified driver behavior as the most influential factor on road safety. Tao, et al. [50] confirmed that traffic violations, vehicle speeding, texting and phone use, intoxication, and sudden lane changes exert a positive influence on the risk of accidents. These risky driving behaviors directly influence the perceived quality of the service [44,46,47,51] and positively influence user satisfaction [42,52,53].

Safety of Transportation Systems
Procedures for safety among public bus drivers, such as using the left lane for picking up passengers, parking at designated points or places in safe areas, and driving assigned routes to pick up and drop off students were significantly positively correlated with perceived safety [52] and influenced satisfaction with the safety of using the public transport services [49].

Safety Policy
Perceived safety and stability while using public transport directly influence passenger loyalty [52] and are factors that influence the perception of the quality of the public transport services [41,45,47,49]. The importance of the perceived quality of services in terms of the perceived safety inside vehicles influences the satisfaction of selecting a mode of public transport [42,43]. Table 1 shows previous studies on the quality of the service and the safety of public transport. There are numerous studies on children's school trips, mainly focusing on demographics [54], socioeconomics [55], barriers to using public transport to school [56][57][58], driver personality [59,60], and environmental awareness [61]. In a study to assess the quality of public transport, [18,[62][63][64] studied the satisfaction and loyalty of the use of tour buses for field trips, on a study of tour drivers in a business context. Ratanavaraha and Jomnonkwao [19] studied the problems of bus drivers. In addition, Zhang, et al. [65] focused on vehicle problems; however, each study had different research questions based on its objectives. This study highlights the perceptions of parents influencing schoolchildren's decisions about the school commute [66], including the service and safety, in order to create strong policy in developing countries [67]. In particular, we focus on the issue of child safety when commuting to school. This study is urgently needed as an effort to advance safe education for children.

Participants and Data Collection
The questionnaire consisted of three main sections, namely, socioeconomic data, school commute environment, and information on the vehicle condition, with a total of 47 questions. For the CFA, the maximum likelihood estimation requires approximately 15 times the number of samples (or 750 samples); [68]. Therefore, data for this study were derived from 750 self-report questionnaires, and the measurement model for the safety management of school travel in Northeast Thailand was analyzed. The first part of the questionnaire pertained to parent information, and second part was student information. Data obtained from the survey are shown in Table 2. The third part consisted of 26 items, which were related to the quality of the public transport services (school trips) for students, which were classified into six groups, namely, Sustainability 2022, 14, 13005 7 of 20 transportation (six items), infrastructure (six items), information (three items), driver safety behavior (four items), transportation safety system (three items), and safety policy (four items). The factors were categorized into six components (i.e., transportation, infrastructure, information, the safe behavior of drivers, the safety of the transportation systems, and the safety policy, as shown in Table 3. Table 3. Description of each indicator.

Code
Variables References Trans Transportation S1 There are services or facilities to use public transport to travel to school.
There is a regular public transport service that connects to the school. [20,69,70] S3 In the area, various types of public transport can be selected for commuting to school (e.g., school buses, vans, and public buses

Data Analysis
First, the study analyzed the descriptive statistics to examine the normal distribution of data using the mean, standard deviation (SD), skewness, and kurtosis. We then used the EFA and principal component analysis (PCA) to group the observed indicators. We introduced the results of the EFA into the CFA to verify the hypothesis. As the composition was complex, this study conducted third-order CFA to gain a deeper understanding of which of the factors were important to transport policy.

Exploratory Factor Analysis (EFA)
This study conducted the EFA using SPSS to find the component variables and PCA to identify the components of a variable. PCA aims to explore the number of components that can be substituted for all variables. In other words, it extracts details from variables into components to reduce the number of variables. For axis rotation, the study used the varimax rotation, which is used to maximize the sum of variances between components, resulting in a clear separation of components and the determination of a highly structured and unique element and leading to a convenient interpretation of components (Little Diamond, 2005). Subcomponents with factor loadings more than 0.5 were considered statistically significant, and each component was composed of no less than three subcomponents [96] with eigen values greater than 1 [97].

Reliability of the Research Instrument
The validity of the research instrument can be verified through construct reliability (CR) and average variance extracted (AVE), which were derived using Equations (1) and (2), respectively: where β denotes factor loading, δ represents error variance, and n pertains to the number of elements considered.

Confirmatory Factor Analysis (CFA)
CFA is a maximum likelihood estimation used to test the consistency of models using empirical data from road safety policy. The recommended criteria for determining the conformity of the hypothetical model to the empirical data for CFA are as follows: the chi-square/df (χ 2 /d f ) should be less than 5 [98]; the comparative fit index (CFI) should be greater than 0.95 Hu and Bentler [99]; the Tucker-Lewis index (TLI) should be greater than 0.8 [100]; the root mean square error of approximation (RMSEA) should less than or equal to 0.08 [101]; and the standardized root mean square residual (SRMR), according to Hu and Bentler [99], should be less than or equal to 0.08. Table 4 provides the details of the model fit indices.

Descriptive Statistics
The basic statistical values used in the analysis to determine the distribution were the mean, SD, skewness, kurtosis, and the correlation between variables. The factors were categorized into six components (i.e., transportation, infrastructure, information, the safe behavior of drivers, the safety of the transportation systems, and the safety policy). Table 5 provides the results. The skewness and kurtosis can represent the normal frequency distribution of the data. The optimum skewness ranges from −2 to 2, and the optimum luminance ranges from −7 to 7. It can be concluded that the data had a normal distribution [103][104][105], as the analysis showed that the skewness of all variables was between −0.33 and 0.13, and the kurtosis ranged from −0.93 to −0.48. All data used in the analysis were within acceptable limits or had a normal distribution and were suitable for further analysis. For the correlation coefficient analysis between variables [106], the correlation coefficient should be between 0.3 and 0.8 to avoid multicollinearity problems [107,108]. The variables in this study were found to have a correlation coefficient of 26 variables between 0.39 and 0.69, as shown in Table 6. Therefore, it can be concluded that the variables used in the study did not show multicollinearity.

Results of the EFA
To determine the composition of a variable (orthogonal axis rotation using varimax rotation and the Kaiser normalization method), the elements identified through the PCA with weights less than 0.5 were excluded [97]. Moreover, the correlation of the indicators for each component must not be less than three items [96]. An element with an eigenvalue greater than 1 indicates that it can be used for factor analysis. EFA is an exploratory technique instead of a reference statistic. Therefore, the results obtained from the EFA were used in the CFA to verify the validity of the hypothesis. Table 6. Results of the correlation coefficient between variables. S1 S2 S3 S4 S5 S6 S7 S8 S9 S10 S11 S12 S13 S14 S15 S16 S17 S18 S19 S20 S21 S22 S23 S24 S25 S26 S1 1.00 S2 0. 63  The basic statistical values used in the analysis to determine the distribution of the data used in the EFA were the mean, SD, skewness, kurtosis, and correlation between variables. The factors were categorized into six components (i.e., transportation, infrastructure, information, the safe behavior of drivers, the safety of the transportation systems, and the safety policy). Table 7 provides the results.

Extraction Results of the EFA
The results of the 26 variables analyzed using the EFA revealed six components (i.e., transportation, infrastructure, information, the safe behavior of drivers, the safety of the transportation systems, and the safety policy), which accounted for 67.652% of the total variance (Table 7). Transportation (component 1; S1−S6) accounted for 45.52% of the total variance, which consisted of six variables. Infrastructure (component 2; S7−S12) accounted for 7.639% of the total variance. The third component was the safe behavior of drivers (component 3; S13−S15) and accounted for 4.348% of the total variance. The safety policy (S16−S19) was the fourth component and accounted for 3.578% of the total variance. The fifth element was information and accounted for 3.433% of the total variance (S20−S22). The last element (the safety of the transportation systems; S23−S26) accounted for 3.137% of the total variance.

Validity and Reliability Tests of the Questionnaire
The study employed the Kaiser-Meyer-Olkin (KMO) and Bartlett's test of sphericity (Table 8) to validate the factor analysis. The KMO values were used to determine the suitability of the sample size, which should not be less than 0.8 [109]. Bartlett's test of sphericity was used to examine the population correlation metrics (identity matrix). The null hypothesis of this test is that the population correlation metric is the identity metric. Therefore, the variables must be classified, such that the significance element of this test should be less than 0.05. In this study, the calculated KMO value of the sample was 0.963, which confirmed that data were suitable for factor analysis. Additionally, a significance level of less than 0.001 was reached using Bartlett's test, which should be less than 0.01. This finding suggests that the correlation metric was not an indicative identity metric. The fact that the variables were related indicates that the data can be used for EFA.

Internal Consistency Reliability
Data from the questionnaire were checked for internal consistency reliability using Cronbach's alpha, which ranges from 0 to 1. The current study obtained Cronbach's alpha values greater than 0.7 [110], and the descriptive statistics of the data from the questionnaire responses in Table 9 indicate the statistical values describing the distribution characteristics, namely, the skewness ranged from −0.20 to 0.02, and the kurtosis ranged from −0.66 to −0.79, indicating normal distribution.

Results of the Third-Order CFA
The results of the third-order CFA, shown in Table 10, demonstrated that each component was positively correlated and statistically significant at the 0.01 level, including the three safety components (the safety of the transportation systems, the safety policy, and the safe behavior of drivers). In other words, these factors are important indicators for determining the safety of public transportation. In addition, the results implied that the satisfaction of parents with transportation, infrastructure, information, and safety are important indicators to consider in determining their perception of the quality of the public transport services used by their children to commute to school. When considering the factor loading of the components, the study found that infrastructure had a factor loading of 0.973, which was the highest value, followed by transport (0.871), information (0.833), and safety (0.816). In terms of the subcomponents of the safety factor, the study observed that safety policy was the most influential factor (0.894), followed by the safety of the transportation system (0.844), and the safe behavior of drivers (0.815). In addition, Table 10 presents the results of the questionnaire validation. For the CR, the factor confidence should be greater than 0.7 [97], and the mean AVE should be greater Sustainability 2022, 14, 13005 13 of 20 than 0.5 [111]. In this study, the CR and AVE values ranged from 0.844 to 0.897 and from 0.570 to 0.654, respectively, which met the recommended values. Figure 1 depicts the results of the CFA on the perceptions of parents regarding the service quality of the public transport for the school trips of their children, which include transportation, infrastructure, information, the safe behavior of drivers, the safety policy, and the safety of the transportation systems. Parental satisfaction, which was related to three safety factors, was classified as a safety factor. The results revealed that the CFA results were consistent with the empirical data (goodness-of-fit statistics: χ 2 = 758.098, df = 276, χ 2 /df = 2.747, CFI = 0.962, TLI = 0.955, SRMR = 0.038, RMSEA = 0.048, * p < 0.05, ** p < 0.01).  Figure 1 depicts the results of the CFA on the perceptions of parents regarding the service quality of the public transport for the school trips of their children, which include transportation, infrastructure, information, the safe behavior of drivers, the safety policy, and the safety of the transportation systems. Parental satisfaction, which was related to three safety factors, was classified as a safety factor. The results revealed that the CFA results were consistent with the empirical data (goodness-of-fit statistics: χ = 758.098, df = 276, χ /df = 2.747, CFI = 0.962, TLI = 0.955, SRMR = 0.038, RMSEA = 0.048, * p < 0.05, ** p < 0.01). Table 11 presents the statistics.

Discussion and Conclusions
This study attempted to confirm the potential indicators affecting the perceptions of the quality of the public transport services by surveying 750 parents whose children use public transport to commute to schools. The EFA results reduced the number of metrics that represented the structure of the data and explained the complex interactions between variables. The study conducted component extraction using PCA, where each component was assigned at least three indicators (eigenvalue > 1), to determine the factors in the parental perception of the public transport service quality of 26 variables. There were six components for the perceptions of parents regarding the quality of the public transport services used by their children to commute to school. The model explained 67.652% of sample variance (i.e., satisfaction with transportation, infrastructure, the safe behavior of drivers, the safety policy, information, and the safety of the transportation systems). The three components of safety (the safe behavior of drivers; the safety policy; and the safety of the transportation systems) can be described as the good safety management of the public transport to school, which was placed in the same order as three other factors, namely, transportation, satisfaction, and infrastructure. The confirmatory factor analysis results revealed that the model was consistent with the empirical data, which reported that the perceptions of parents regarding the quality of the public transport services used by children to commute to school consisted of four components, namely, transport satisfaction, infrastructure, information, and safety, where the safety component was composed of three subcomponents, namely, the safe behavior of drivers, the safety policy, and the safety of the transportation systems. This result was consistent with those of previous study that used the following factors:

Transportation
The results from the CFA confirmed that the transportation factor affected perceptions of the use of public transport for commuting to schools, including:

•
The availability of services or facilities to use public transport for commuting to school, consistent with the findings of de Oña, Estévez and de Oña [20], Atombo and Dzigbordi Wemegah [69], Yao, Xu, Zhang and Li [70]. • There should be a public transportation system that students can use to travel to or near the school grounds, which was in line with the studies of de Oña, Estévez and de Oña [20], Atombo and Dzigbordi Wemegah [69], Yao, Xu, Zhang and Li [70]. • There should be more than one type of public transport service in the area, allowing students to choose to use them as appropriate, which was in line with the studies of de Oña, Estévez and de Oña [20], Atombo

Infrastructure
According to the CFA results, the infrastructure factor affected the perceptions of using public transport for commuting to school, including: • Public transport must be easily accessible, having a drop-off point close to the origin and destination, or students must have access to public transport on their own, consistent with the findings of de Oña, Estévez and de Oña [20], Yao, Xu, Zhang and Li [70]. • There should be clearly visible signs of the public transport system for users, to indicate the pick-up point location and reduce unnecessary stops along the way, which was in line with the studies of Ren, Jin and Wu [77], Taplin and Sun [78].

•
Public transport systems must have clear service routes, enabling students and parents to use them in planning their trips, and providing information to other users of the service, in line with the studies of Ren, Jin and Wu [77], Taplin

Information
The results of the analysis showed that the information factor influenced perceptions of the use of public transport, which included: •

Safety
Due to the large number of safety variables, the study was analyzed using third-order confirmatory factor analysis for the strength of the model. The analyses were divided into three main subfactors, and the CFA analysis discovered that: (1) Driver safety behavior influences the perceptions of using public transportation for commuting to school. The important factors included drivers using the appropriate speed, according to the law, and safely, which was consistent with the studies of Mokarami, Alizadeh, Rahimi Pordanjani and Varmazyar [14], Karimi, Aghabayk and Moridpour [83], Davey, Wishart, Freeman and Watson [84], Ammar, Jalmoud, Boushehri and Fakhro [85]. It also included drivers not engaging in risky behaviors such as sudden lane changes (which corresponded to the studies of Karimi, Aghabayk and Moridpour [83], Varmazyar, Mortazavi, Hajizadeh and Arghami [89], Hendrix and Kennedy [90]). Further, drivers should not use phones or other communication devices while driving (which was consistent with Ammar, Jal-moud, Boushehri and Fakhro [85], Valero-Mora, Zacares, Sánchez-García, Tormo-Lancero and Faus [86], Chee, Irwin, Bennett and Carrigan [87]). Finally, drivers must not behave inappropriately as employees, such as smoking or using intoxicants while driving, consistent with the Ratanavaraha, Jomnonkwao, Khampirat, Watthanaklang and Iamtrakul [18], Ratanavaraha and Jomnonkwao [19], Chang and Yeh [88] studies. (2) The safety of the transportation systems influenced the perceptions, consisting of public transport using the left lane for pick-up and drop-off, to increase the safety of passengers. (Thailand uses the left lane to drive), which was consistent with the studies of Mokarami, Alizadeh, Rahimi Pordanjani and Varmazyar [14], Huang, Lin and Wang [91]. Agreeing with the Mokarami, Alizadeh, Rahimi Pordanjani and Varmazyar [14], Huang, Lin and Wang [91] findings, public buses must stop in a safe spot when picking up or dropping off passengers. The driver must park on the left side and not drive off the road unless there is an emergency, as this will cause concern to passengers. (consistent with the results of Ammar, Jalmoud, Boushehri and Fakhro [85], Hendrix and Kennedy [90]). (3) The safety policy also influenced the perceptions; these factors include staff being required to check student orders before departure to check passenger safety (as in Salam, Muley and Kharbeche [71], d'Ovidio, Leogrande, Mancarella, Schinzano and Viola [72], Agyeman and Cheng [73], Hendrix and Kennedy [90]). Public transport vehicles must be equipped with emergency accident prevention equipment such as fire extinguishers (for large vehicles) and guardrails to prevent falling children (in small vehicles), which was consistent with the studies of Ratanavaraha, Jomnonkwao, Khampirat, Watthanaklang and Iamtrakul [18], Jomnonkwao and Ratanavaraha [92], Davis and Abulhassan [93], Olympia, Weber, Brady and Ho [94]. There should be a support or anti-theft device, which corresponded to the study of de Oña, Estévez and de Oña [20], Atombo and Dzigbordi Wemegah [69], Yao, Xu, Zhang and Li [70], d'Ovidio, Leogrande, Mancarella, Schinzano and Viola [72], and public transportation must be safe from road accidents, which was consistent with the findings of de Oña, Estévez and de Oña [20], Atombo and Dzigbordi Wemegah [69], Sukhov, Lättman, Olsson, Friman and Fujii [82], Deb, Ali Ahmed and Das [95].

Limitations and Suggestions for Future Studies
Although this study fulfilled the objectives for a study on the parental perceptions of the quality of children's public transport services on school trips, this study did not enter the heterogeneity of the sample into the analysis. Therefore, future studies considering heterogeneity may be able to better describe the results of the analysis. In addition, analyzing models with SEM can lead to more accurate and complex robust structures.