Pedestrians in Iran: Determinants of Unsafe Trac Behaviors of Pedestrians

Background: Pedestrians are among the most vulnerable groups in trac accidents. This study investigates the determinants of trac behaviors of pedestrians. Methods: In this cross-sectional study, the behavior of 600 pedestrians in the city of Urmia, northwestern Iran, was evaluated from May to October 2018, using a pedestrian trac behavior questionnaire consisting of ve dimensions: trac violations, trac distraction, adherence to trac rules, aggressive behavior, and positive behaviors. Cluster sampling was conducted among 30% of the health centers in each district of the city. Thereafter, individuals were randomly selected from each center depending on the covered population. Results: Although 17% of the pedestrians demonstrated safe trac behaviors, over 85% were distracted, more than 80% did not comply with trac laws, the majority violated rules and displayed no positive trac behaviors, and half of the participants were aggressive. Age, gender, marital status, education level, daily walking rate, and transportation mode were signicant predictors of trac behavior. Conclusion: Most pedestrians demonstrate unsafe trac behaviors. The youth, those who were single, those with lower education, and those with less walking rate showcased higher unsafe behaviors. Training and intervention programs should be implemented to improve all domains of pedestrian trac behaviors factoring in age and other predictors. The total explained variance for PTB was 12%, and R square was 0.06 for compliance with laws, 0.08 for trac violation, 0.05 for positive behavior, 0.09 for trac distraction, and 0.07 for aggressive behavior. had a signicant impact on the range of violent and aggressive behaviors. Adherence to trac rules was signicantly lower among men than among women, and trac violation and demonstrating aggressive behaviors were signicantly higher among men than among women. Education had a signicant impact on the domains of aggressive behavior and trac violation. The transportation mode had a positive impact on the domains of trac distraction and trac violation. We further found that had a signicant impact on trac violation, trac distraction, and aggressive behaviors.


Background
Injuries caused by tra c accidents are the eighth leading cause of death across age groups (1). Pedestrian safety is one of the most important (2) as well as neglected (3) issues in the world. At the international level, pedestrians constitute about 22% of the total road casualties and, in some countries, this ratio even reaches two-third of their total population. Millions of pedestrians are injured in tra c accidents while walking, some of whom permanently remain disabled and paralyzed (4). Pedestrian deaths, in general, are higher in low-and middle-income countries compared with high-income countries (45% road deaths in low-income countries, 29% in middle-income countries, and 18% in high-income countries) (5,6). The Eastern Mediterranean region is an exemption where high-income countries still have high pedestrian death rates (5).
There is no global estimate of the economic impact of pedestrian accidents, but it has been estimated that between 1 and 2% of the GDP is being consumed for road accidents (1). Survivors of accidents along with their families, friends, and others often suffer from adverse social, physical, and psychological effects. Road tra c deaths in Iran are among the highest in the world and the Eastern Mediterranean region, and in line with those in African countries (1). Pedestrians are vulnerable groups and there are no effective laws to control their behaviors. Majority of the accidents occur because pedestrians do not comply with tra c laws (7). Two-third of the U.S. pedestrians enter the streets paying no attention to vehicles and 16% cross the street without paying attention to vehicles. Among those who notice vehicles approaching, 40.6% stopped, 11.4% return to let the vehicles pass by, and 31.9% rush to cross the street. Some are so overwhelmed with talking on their phone that they forget they are crossing the street (2).
Various studies have shown that by observing pedestrians who incorrectly cross the road, many people are likely to follow them carelessly (8)(9)(10). Some works have pointed out factors such as distraction (mobile phones) (11), pedestrian perception of risk and its relationship with highrisk behaviors (12,13), personality characteristics, age, and gender (14,15), alcohol use, and not seeing pedestrians on the roads (16,17) as high-risk pedestrian behaviors. In Iran, over 39% of the deaths caused by tra c involve pedestrians (18), but to the best of our knowledge, there is no study that has investigated the status of pedestrian tra c behaviors (PTB) and their different domains in Iran. This study aims to investigate PTB in different domains and to determine the predictors of unsafe behavior. We add to the literature by focusing on the determinants of the ve tra c behaviors: tra c violations, tra c distraction, adherence to tra c rules, aggressive behavior, and positive behaviors.

Sample Design
This descriptive cross-sectional study was conducted among 600 pedestrians in the city of Urmia, northwestern Iran, from May to October 2018. Cluster sampling method was applied. Given that in Iran, family les and complete information of all households are kept in health centers, we chose to collect data in health centers. The health centers of each clustert were randomly selected. Then, from each center, depending on the population covered, the participants were randomly selected with respect to the inclusion and exclusion criteria.
The inclusion criteria were (a) being 18 years old or above, (b) willing to participate in the study, and (c) being capable of standing and walking. The exclusion criteria were having a history of severe mental illness, depression, Alzheimer's, dementia, restrictive musculoskeletal disorders, neurological de cits (stroke), Parkinson's and paralytic disease, acute heart diseases (acute myocardial infarction, uncontrolled hypertension), or severe hearing and visual impairment. After selecting the respondents, they were invited to visit the health centers to participate in the study. All participants completed the informed consent form. Based on Jalilian et al.'s (19) work and considering mean and standard deviation of pedestrian behaviors, cluster sampling, effect size, as well as rate of loss, the sample size was estimated at 600. Data in this study were collected by means of The Pedestrian Behavior Questionnaire (PBQ), which had been designed by the Road Tra c Injury Research Center, Tabriz University of Medical Sciences. The instrument proved to be valid and reliable (20). The instrument consisted of 29 items measuring pedestrian behaviors across the following ve dimensions, using a ve-point scale ranging from 1 to 5 (1=never, 2=rarely, 3=sometimes, 4=often; and 5=always): (1) adherence to tra c rules (7 items) (e.g., I cross the street after the vehicles are fully stopped); (2) tra c violations (10 items, reverse scored) (e.g., I don't use the pedestrian bridge because most people don't); (3) positive behaviors (2 items) (e.g., I let the vehicle pass even if I have the priority right); (4) tra c distraction (4 items, reverse scored) (e.g., I use my mobile phone while crossing the street); and (5) aggressive behaviors (2 items, reverse scored) (e.g., If I get angry at the behavior of a driver, I would kick or punch the car). Total PTB score is calculated as the sum of scores received in each of the ve domains. The higher the score, the safer the behavior. Cronbach's alpha of PTB was 0.83. The number of items in each domain and possible ranges are shown in Table 2. Raw score of each domain was calculated as sum of scores of items of that domain. Standardized score for each domain was considered as mean score of items in each domain. PTB and each domain of behavior were categorized into three groups, poor (≥95), moderate (96-124), and good (≤125), based on the mean and SD (see Table 2).

Ethical Approval
This study was conducted in accordance with the ethical standards of the 1964 Helsinki declaration. Ethical approval was provided by the ethical committee of Tabriz University of Medical Sciences vide reference number IR.TBZMED.REC.1397.969.

Analytical Strategy
The analyses were carried out using SPSS (version 23). Mean and standard deviations are provided for continuous variables and number and percentage for categorical variables. The normality of distribution for the data was checked by applying the Kolmogorov-Smirnov test. To assess the relationship of the total PTB score with the demographic variables, a univariate regression was applied. Multivariate regressions were performed to examine age, gender, education, marital status, educational level, daily walking rate, and transportation mode as independent variables to determine a set of best predictors for the ve domains as the dependent variable. Hotteling's Trace method was used to calculate the p-value and evaluate the impact of the variables.

Results
As shown in Table 1, the mean age of the participants was 36.69 (SD=10.69) years. The majority of the participants were women (60.2%) and married people (70.5%). The highest level of education completed by most participants was less than a high-school degree (52%). In total, 43.8% of the participants walked less than 30 minutes per day. Over 60% of the participants used a personal car for transportation. Table 1 includes the ve different domains together with the other study variables. Univariate regression analyses revealed that tra c behaviors were better among the elderly compared to the other age groups, among females compared to males, and among the married compared to others.
As the level of education increased, the tra c behaviors became safer. The behaviors of bicyclists and motorcyclists were less safe than those using other transportation modes. The people with longer daily walking rate showcased safer behaviors.   Table 2 presents an overview of PTB and the different domains in terms of demographic variables. As shown in the table, tra c behaviors were good among only 17% pedestrians. Over 85% pedestrians experienced tra c distraction and over 80% of them did not adhere to tra c laws. Most of them showcased tra c violations and demonstrated no positive tra c behaviors, and half of participants even demonstrated aggressive behavior. In the present study, the score in the "aggressive behavior" domain of the studied pedestrians was better than that in the other domains, while the score in the "adherence to tra c rules" domain was lower than that in the other domains.
Results of a series of univariate regression analyses demonstrated that all the study variables (including gender, marital status, education level, daily walking rate, transportation type, and age) were signi cantly associated to PTB. Among those aged below 30 years, the tra c behaviors were better than that for those aged more than 30 years. More particularly, in the age group of the elderly, PTB was 10.7 points higher than that for the youth group. Total score of men was 3.83 points lower than that of women and those who were single had more unsafe behaviors than the others and their score was 8.57 points lower than the rest. Married people displayed safer behaviors compared to those who were single and others. The higher the level of education, the more signi cantly improved were the tra c behaviors.
Multivariate regression results showed that, in the different domains of behaviors, several variables were signi cantly associated with PTB. The total explained variance for PTB was 12%, and R square was 0.06 for compliance with laws, 0.08 for tra c violation, 0.05 for positive behavior, 0.09 for tra c distraction, and 0.07 for aggressive behavior. Gender had a signi cant impact on the range of violent and aggressive behaviors. Adherence to tra c rules was signi cantly lower among men than among women, and tra c violation and demonstrating aggressive behaviors were signi cantly higher among men than among women. Education had a signi cant impact on the domains of aggressive behavior and tra c violation. The transportation mode had a positive impact on the domains of tra c distraction and tra c violation. We further found that age had a signi cant impact on tra c violation, tra c distraction, and aggressive behaviors.

Discussion
The purpose of this study was to investigate the tra c behaviors of pedestrians in the city of Urmia and to determine their most important predictors. In this study, less than 20% of the pedestrians demonstrated safe tra c behaviors. Over 85% of the pedestrians were distracted and over 80% of them did not adhere to tra c rules. The majority demonstrated tra c violations and no positive tra c behaviors and half of them were aggressive. This implies dangerous and high-risk behavior among pedestrians in Urmia. Contrary to the present study, other studies have shown that the majority of the pedestrians had positive behaviors and were less aggressive (2,21). Although the majority of the participants in the present study were distracted and the frequency of aggressive behaviors was lower than other domains, almost half of the participants had aggressive behaviors, which is higher than other studies. In a study conducted in the U.S., 12% of the participants said they were aggressive and 15% expressed they were slowly crossing the street to annoy drivers (2). In the present study, the tra c behaviors were inappropriate in all domains. Therefore, it is important to provide the necessary training to pedestrians in all domains of behavior in order to change the overall tra c behavior (1, 7). One study showed that a distracted pedestrian can cause aggression among others, such that a pedestrian who is both aggressive and distracted suffers more in accidents (22). Therefore, due to the inappropriate behavior in different domains as well as the unsafe tra c behavior of most of the studied pedestrians, it is necessary to intervene in all domains of tra c behaviors to reduce pedestrian accidents.
To reduce intervention costs, it is essential to perform various training interventions, make changes in the physical environment, and increase policies to promote positive pedestrian behaviors. One way to do this is by conducting training workshops in the selected neighborhoods on pedestrian and cyclist's safety (22), designing new pavements with different performance for pedestrian safety (23), providing re ective products (such as cloth strips and LED silicone band) and training brochures (24), installing pedestrian signboards in the middle of two-way lanes (YTPCD) (25), increasing compliance with priority rights of pedestrians by drivers and stopping to let them pass (26), and increasing positive behaviors and mutual respect alongside the improvement in Iranian infrastructure and culture.
Given the high distraction rate in the present study, it seems necessary to implement training programs. For instance, in several cities the physical environment has been changed by establishing alert systems, special routes for people who use mobile phones, pedestrian islands, etc. in the design of streets (Santa Monica, California, United States, 2016, Jersey City, New Jersey, United States, 2017) or designs such as Pedestrian Friendly Streets (Honolulu, United States, 2015). To increase regulation in the society, it is essential to begin tra c law training from childhood and continue it over time (1).
According to the ndings of this study, all the demographic variables were signi cantly associated with tra c behaviors. Age was one of the most important predictors of PTB, so that with aging, people demonstrated higher levels of safe behaviors. This nding is similar to the result of other studies (2,21). In the present study, the PTB score of the elderly was 10.7 points higher compared to that of those aged under 30 years. In examining behavioral dimensions, it was also observed that age had a signi cant impact on tra c violation, tra c distraction, and aggressive behavior.
This was similar to another study (27) wherein age was found to be signi cantly negatively associated with aggression and violation. It could be said that the higher experience of the elderly could lead them to have safer behaviors. Moreover, it could be argued that adolescents are more likely to use mobile phones and other distractive devices that could increase their distraction. A study showed that almost 100% of the Americans aged 18-29 years had mobile phones and about 92% had smartphones (28). Mobile use is one of the important distractors and environmental and training intervention programs are required in this regard. Given that violation was high among the youth, the training programs must begin at an early age and continue thereafter.
This study showed that gender was a predictor of pedestrian behaviors and women showcased safer behaviors than men. According to the World Health Organization (1), the tra c accident death rate in Iran was 15932 in 2016, of which 78% included to men (1). A research carried out in the United States showed that 70% of the pedestrian deaths was related to men (mortality rate of 2.19 per 100,000 people compared with 0.91 in women). In Mexico, the pedestrian mortality rate among men (10.6% per 100,000 people) was higher than that among women (4% per 100,000 people) (29,30). In the present study, the men and those who were single were less adhering to tra c rules, and they demonstrated higher violation and higher aggression. A greater emphasis on men is required in the interventions for safe tra c behaviors.
Results of the present study showed, with an increase in education level, the tra c behaviors of the pedestrians signi cantly improved, which was similar to the ndings of other studies (3,31). Higher education had a signi cant impact, especially on the dimensions of adherence to tra c rules and aggressive behaviors. Studies have shown that increased level of education also affects the acquisition of other healthy behaviors, such as physical activity (32,33). It can be argued that, with increased education level, the perception of various issues is increased, so people can better understand safe behaviors.
This study showed that, in individuals with more walking rate, adherence to tra c rules was signi cantly better. Also, higher experience of walking and observing behaviors of others in the environment makes people practically understand the laws. Studies have also shown that seeing the behavior of others in the environment leads to the same behaviors among the individuals in Iran, the rate of death and injury from tra c accidents is high. Therefore, it is essential to examine the behavior of users and determine the predictors for implementing interventional programs with the aim of reducing the burden of tra c accidents. Thus, one of the strengths of this study could be in terms of investigating the behavior of pedestrians as vulnerable users, which was carried out in one of the large cities of northwestern Iran. Urmia is one of the cities in which different ethnic groups of Turks and Kurds with different cultures are living, and the cultural context of the region can affect the behavior of the people in that domain. Moreover, the questionnaire applied to measure the behavior was localized for Iran. The questionnaire had been developed and validated by the Tra c Research Center of Iran. However, the present study also has some limitations, including tra c accident history of the participants or their family and friends, language, and ethnicity were not assessed, which could affect the tra c behavior. Another limitation is the use of self-reports in this study which may have resulted in biased answers. Therefore, it is suggested to have observational and behavioral measures to better tap the behaviors of pedestrians. Furthermore, other factors that could predict behaviors, such as history of tra c accidents, must be considered.

Conclusions
The present study found that the behaviors of the Iranian pedestrians were rather unsafe. The youth, those who were single, those with lower education, those with less walking rate, as well as bicyclists and motorcyclists demonstrated more unsafe behaviors. It is essential to have interventional training programs to improve all domains of tra c behaviors with respect to age and other predictors.

Declarations
Availability of data and materials All relevant data are available within the manuscript. In case of need, the data that support the ndings of this study are available from the corresponding author on reasonable request. Declaration of Competing Interest: The authors declare that they have no competing interests.