Modeling the Influence of Smartphone Distraction and Pedestrian Characteristics on Pedestrian Road Crossing Behavior

Abstract

With the increase in the use of technology, smartphone distraction has become a significant pedestrian safety issue. Research on pedestrian smartphone distraction has been conducted mainly in western countries, and there is a scarcity of research in low- and middle-income countries (LMICs). This study is the first attempt to investigate the effect of smartphone distraction on pedestrians’ road crossing behavior in India. We conducted a survey of pedestrians, with 560 respondents (63.4% males), about their demographics and walking characteristics, pedestrian behavior (violations, errors, lapses, aggressive behaviors, and positive behaviors), and smartphone distraction. The associations of smartphone distraction and pedestrian characteristics with pedestrians’ road crossing behavior were identified by developing structural equation models (SEMs). The strongest influence of smartphone distraction was found to be on violations (factor loading = 0.50), followed by errors (factor loading = 0.44), lapses (factor loading = 0.41), and aggressive behaviors (factor loading = 0.37). No significant effect of smartphone distraction was observed on the positive behaviors. Male pedestrians were more likely to commit violations, errors, lapses, and aggressive behaviors, whereas female pedestrians showed more positive behaviors. Older pedestrians showed more positive behavior than young pedestrians. Pedestrians who walk at high speeds tend to engage in a high number of violations. Pedestrians with a crash history were more likely to commit violations and show aggressive behaviors than those with no prior crash experience. The study findings can aid policymakers in understanding the smartphone distraction habits of pedestrians, which are required to design behavior-based road safety interventions.

Keywords

pedestrians human factors cell-phone use distraction walkability

In recent years, walking has been promoted by researchers and governments around the world as it has positive impacts on physical, psychological, social, economic, and environmental health ( 1 , 2 ). Walking is a basic mode of transportation for making trips ( 3 ). Even though walking has several benefits, pedestrians are the most vulnerable group reported in road traffic fatalities ( 4 , 5 ). The recent global road safety report states that 23% of road deaths are comprised by pedestrians ( 6 ). Worldwide, observational studies have reported that pedestrians are often found to commit road safety violations, especially during road crossings at midblock sections and intersections ( 7 – 12 ).

Mobile phones have become an integral component of everyday life. With new technologies penetrating the consumer market every day, the use of mobile phones has significantly increased over the years. Considering this growth, technological distraction among pedestrians has emerged as an interesting area for road safety researchers. Many times, pedestrians have been observed to be engaged with their mobile phones during the task of road crossing ( 13 ). The pedestrians who are found to use mobile phones during road crossing are also found to have slow walking speeds, which makes them more vulnerable to being involved in road traffic crashes ( 14 ). A recent systematic review suggested that the act of mobile phone use during road crossing has been observed in 12% to 45% of pedestrians, which is quite alarming ( 15 ).

The detrimental impact of smartphone distraction on pedestrian safety has been well established in laboratory experiments ( 16 ), observational studies ( 17 ), crash-based investigations ( 18 ), as well as survey-based research investigating pedestrian psychology ( 19 ). In a recent study, Liu et al. ( 20 ) modeled the effects of visual and auditory mobile phone distractions on the reaction time of pedestrians in response to green signals. They found that Chinese pedestrians responded more quickly than Canadian pedestrians, and the pedestrian reaction was faster at intersections than midblock crosswalks. Visual distractions resulted in a 50% increment in the reaction time of pedestrians, while this increment was 67% in the case of auditory distractions ( 20 ). A new term has been coined in the recent literature, where the pedestrians engaged in their mobile phones are termed as “smartphone zombies,” or “smombies” ( 21 ). The type of mobile phone distraction can be cognitive, auditory, visual, or physical ( 22 ). Different types of distraction activities lead to varying levels of attention deficit among pedestrians ( 20 ). For instance, texting and web surfing have resulted in a higher unsafe road crossing behavior among pedestrians compared with talking and listening to music ( 23 ). Lennon et al. ( 24 ) reported that pedestrians in the age group of 18 to 30 years have a significantly higher likelihood of using a mobile phone during road crossing compared with older pedestrians. Lin and Huang ( 25 ) found different gaze scanning patterns and event response performances among the pedestrians involved in various distraction activities. With an increase in mental workload, situational awareness among pedestrians diminished ( 25 ). These findings highlight the importance of understanding the behavioral aspects of pedestrian distraction in the fast-growing technological world ( 26 ).

The literature on pedestrian technological distraction is far less than the plethora of study on driver distraction ( 27 , 28 ). The studies on pedestrian smartphone distraction have been conducted mainly in developed countries, and there is a scarcity of research from the perspective of low- and middle-income countries (LMICs)/developing countries. Traffic conditions vary widely between the developing and developed countries. Where developed countries have lane-based homogeneous traffic with dedicated lanes for traffic movement, developing nations have no-lane-based heterogeneous traffic movements, that is, the vehicles do not strictly follow lane-based discipline and drive through the gaps between the lead vehicles. Moreover, compliance with the law and adherence to traffic rules by road users are more stringent in developed countries than in developing countries. Additionally, developed countries are well equipped with pedestrian safety infrastructure/facilities, adequate pedestrian phase/signal timing for road crossing, and so forth, which is not as prevalent in a developing nation like India. A recent study on cross-cultural comparison of pedestrian behaviors reported that pedestrians in developing countries have higher tendencies to perform risky road crossing behavior than those in developed countries ( 29 ). As more than 90% of road fatalities occur in LMICs, there is a need to conduct detailed examinations on pedestrian distraction ( 6 ). Further, much of the research on pedestrian distraction has been done through traffic observations and laboratory experiments, and relatively few studies have utilized self-reported measures for understanding pedestrian behavior. In addition, the Pedestrian Behavior Questionnaire has been widely used to study the risky road crossing behavior of pedestrians, and the association of self-reported pedestrian behavior with mobile phone distraction is scarcely studied. Therefore, the present study used a self-reporting approach to examine the behavioral impairment among Indian pedestrians when using their mobile phones during road crossing. This study aims to test two major hypotheses. First, it can be hypothesized that smartphone distraction leads to a higher frequency of violations, lapses, errors, and aggressive behaviors among pedestrians. The second hypothesis is that pedestrians involved in higher levels of smartphone distraction tend to show lower positive behaviors.

In this context, the objectives of the present study are as follows:

a. To investigate the influence of self-reported smartphone distraction frequencies on the road crossing behavior of Indian pedestrians. Road crossing behavior has been categorized into five aspects: violations, errors, lapses, aggressive behaviors, and positive behaviors ( 29 , 30 ).

b. To evaluate the associations of the demographics and walking characteristics (age, gender, crash history, walking duration, walking speed, and quality of walking environment) with the road crossing behavior of pedestrians.

This paper is structured as follows: the detailed discussions on the Pedestrian Behavior Questionnaire, smartphone distraction, and the role of pedestrian characteristics along with the data collection procedure are provided in the survey design section. Further, the sample characteristics, descriptive statistics obtained from the survey, and data analysis technique are described. Next, the detailed data analysis, model results, and their interpretations are illustrated. Subsequently, the study findings are discussed in the light of past research along with practical implications of the findings. At the end, the limitations and future research scope of the study are highlighted, and the conclusions are outlined.

Methods

Survey Design

Various versions of the Pedestrian Behavior Questionnaire (PBQ) have been utilized in the literature, which have been adapted from the Driver Behavior Questionnaire (DBQ) given by Reason et al. ( 31 ). The initial structure of DBQ consisted of two factors: violations and mistakes, where the violations represented intentional behavior while mistakes depicted the unintentional errors. Parker et al. ( 32 ) further categorized driver behavior into three factors: violations (intentional behavior), errors (unintentional mistake), and lapses in attention. Further, Lawton et al. ( 33 ) added the fourth component of aggressive behavior, and Özkan and Lajunen ( 34 ) included the fifth component of positive behavior. Over the years, the PBQ has been used to understand pedestrian behavior in various countries such as the United States of America ( 30 ), Serbia ( 35 ), Iran ( 36 ), Australia ( 37 ), Spain ( 38 ), France ( 39 ), Turkey ( 40 ), Chile ( 41 ), and the United Kingdom ( 42 ). Deb et al. ( 30 ) introduced a short version of the 20-item PBQ with five items (violations, errors, lapses, aggressive behaviors, and positive behaviors) developed from the lengthy version of the 50-item PBQ and demonstrated its efficacy in examining pedestrian behavior at crosswalks. McIlroy et al. ( 29 ) used this 20-item PBQ to analyze the differences in pedestrian behavior among six distinct countries: China, the United Kingdom, Bangladesh, Vietnam, Kenya, and Thailand. A summary of previous studies investigating pedestrian behavior using PBQ is shown in Table 1.

Table 1.

Summary of Past Studies Investigating Pedestrian Behavior Using the Pedestrian Behavior Questionnaire (PBQ)

Study	Country	Factors	Items	Likert scale	Sample size
Díaz ( 43 )	Chile	3	Violations, errors, lapses	5-point	146
Elliott et al. ( 44 )	United Kingdom	3	Unsafe road crossing, dangerous playing on the road, planned protective behavior	5-point	2,433 children
Granié et al. ( 39 )	France	4	Transgressions (including violations and errors), lapses, aggressive behaviors, positive behaviors	6-point	343
Nordfjærn, and Şimşekoğlu ( 40 )	Turkey	3	Transgressions (including violations and errors), violations, aggressive behavior	5-point	289
Antić et al. ( 35 )	Serbia	5	Violations, errors, lapses, aggressive behaviors, positive behaviors	6-point	415
Qu et al. ( 45 )	China	4	Transgressions (including violations and errors), lapses, aggressive behaviors, positive behaviors	5-point	283
Deb et al. ( 30 )	USA	5	Violations, errors, lapses, aggressive behaviors, positive behaviors	6-point	425
McIlroy et al. ( 29 )	China, United Kingdom, Bangladesh, Vietnam, Kenya, and Thailand	5	Violations, errors, lapses, aggressive behaviors, positive behaviors	6-point	3,423 (across six countries)
O’Hern et al. ( 37 )	Australia	4	Violations, errors, aggression, technology	6-point	968
Useche et al. ( 38 )	Spain	3	Violations, errors, positive behaviors	5-point	1,070
Esmaili et al. ( 36 )	Iran	4	Transgressions (including violations and errors), lapses, aggressive behaviors, positive behaviors	6-point	520

In this study, a questionnaire was designed with different sections on pedestrian demographics, walking characteristics, smartphone distraction frequencies, and pedestrian road crossing behavior. The Institutional Review Board (IRB) of the Indian Institute of Technology Bombay approved the study protocol (Reference: IRB-2021-021). Pedestrian road crossing behavior was evaluated using the 20-item PBQ, which has been widely used in previous studies on pedestrian behavior ( 29 , 30 , 39 ). The PBQ consisted of five constructs: violations, errors, lapses, aggressive behaviors, and positive behaviors ( 30 ). All five constructs consisted of five items each. The pedestrians had to rate their respective behavior on a five-point Likert scale where “1” represented “never or very infrequently” and “5” signified “always or very frequently.” Further, smartphone distraction frequencies were captured using the 10-item instrument covering the various distracting activities, adapted from Young et al. ( 46 ).

Along with smartphone distraction, pedestrian characteristics have also been found to play an essential role in influencing the road crossing behavior of pedestrians. Previous studies have examined the role of various contributing factors such as pedestrians’ age ( 17 ), gender ( 47 ), walking duration ( 48 ), quality of walking environment ( 49 ), and walking speed ( 16 ) on their road crossing behavior. In a recent study, older pedestrians showed safer road crossing behavior than younger pedestrians ( 50 ). Further, male pedestrians were found to have 4.1 times higher odds of displaying unsafe road crossing behavior at unsignalized intersections compared with female pedestrians ( 17 ). Therefore, in addition to the PBQ items, pedestrians’ age, gender, crash history, and walking characteristics, such as the walking duration on a typical day, quality of the walking environment, and self-reported walking speed, were also recorded in the questionnaire. For the crash history, the respondents were asked whether they had any experience of a prior crash as a pedestrian or not. Responses on the quality of the walking environment were on a five-point Likert scale where “1” indicated “very poor” and “5” indicated “very good.” The description of each variable asked in the questionnaire is provided in Table 2.

Table 2.

Demographic and Walking Characteristics of Respondents

Variable	Variable description	Categories/levels	Frequency	Percentage (%)
Age	What is the age of the respondent (in years)?	1. 18 to 25 years	310	55.36
		2. 26 to 35 years	172	30.71
		3. 36 to 50 years	59	10.54
		4. More than 50 years	19	3.39
Gender	What is the gender of the respondent?	0. Female	205	36.61
Gender	What is the gender of the respondent?	1. Male	355	63.39
Crash history	Has the respondent been involved in a crash in the past as a pedestrian?	0. No	413	73.75
Crash history		1. Yes	147	26.25
Walking duration on a typical day	How much does the respondent walk on a typical day (in minutes)?	1. Less than 15 min	133	23.75
		2. 15 to 30 min	220	39.29
		3. 30 to 45 min	129	23.04
		4. 45 to 60 min	50	8.93
		5. More than 60 min	28	5.00
Quality of the walking environment	How does the respondent rate the quality of the surrounding environment while walking?	1. Very poor	47	8.39
		2. Poor	130	23.21
		3. Moderate	256	45.71
		4. Good	118	21.07
		5. Very good	9	1.61
Self-reported walking speed	How does the respondent rate their individual walking speed?	1. Very slow	4	0.71
		2. Slow	29	5.18
		3. Moderate	180	32.14
		4. Fast	273	48.75
		5. Very fast	74	13.21

The survey was circulated on online platforms and social media groups where the participation was restricted to Indian pedestrians with the age of 18 years and above. The data collection was conducted between September and November 2021. The average survey duration was about 15 min. Participation was voluntary and anonymous, and no monetary rewards were provided. Only complete responses were recorded as it was mandatory for all the questions to be answered.

Sample

In this study, the target sample size was calculated using Cochran’s formula ( 51 ), shown below:

n = \frac{Z^{2} pq}{d^{2}} = \frac{{1.96}^{2} \times 0.5 \times 0, 5}{{0.05}^{2}} = 384

(1)

where p represents the target population (0.5 for a very large population), q = 1 −p, d is the error margin (5%), and considering 95% confidence interval, equivalent Z = 1.96. Accordingly, the responses collected in the survey were more than the target sample size.

In total, 592 participants completed the survey. Out of these, 32 responses were filtered because of incorrect responses provided in either of the two check questions included in the survey. The remaining 560 participants’ responses were considered for the final analysis. Table 2 shows the demographic and walking characteristics of the final sample: 63.4% of the participants were males, and 36.6% were females. About 55.4% of the participants were in the age group of 18 to 25 years. This indicates that the majority of the participation came from younger people as they are more likely to be technology-enabled than older people.

Data Analysis and Results

STATA MP-15 software was used to conduct the statistical analysis in the present study. The preliminary analysis included estimating the reliability/internal consistency of all the scales used in the study. For this purpose, Cronbach’s alpha was obtained for each scale, with values more than 0.7 being considered accepted ( 52 ). Further, two structural equation models (SEMs) were developed. The first SEM model aimed to identify the associations of smartphone distraction frequencies with the five constructs of pedestrian road crossing behavior. The second SEM model was developed to investigate the relationships of pedestrians’ demographic and walking characteristics with the five constructs of PBQ.

Descriptive Statistics

Table 3 presents the descriptive statistics (percentage distribution of responses, mean and standard deviation) of all the 20 behavioral items of the PBQ responded to by the pedestrians. The reliability coefficients (Cronbach’s α) of each of the five subscales are as follows: violations (0.813), errors (0.773), lapses (0.910), aggressive behaviors (0.906), and positive behaviors (0.710), indicating that these subscales were reliable ( 19 ). Pedestrians rated higher frequencies in showing positive behaviors during road crossing (e.g., thanking a driver for stopping and allowing the pedestrian to cross). The aggressive behaviors were found to have the lowest frequency ratings by the pedestrians (Table 3). The ratings of violations, errors, and lapses lay in between these two extremes. Similar reporting was also made by American pedestrians in a previous study ( 30 ).

Table 3.

Descriptive Statistics of Responses to Pedestrian Behavior Questionnaire (PBQ)

Pedestrian behavior items	Responses in % (1: Never; 5: Always)					Mean (standard deviation)
Pedestrian behavior items	1	2	3	4	5	Mean (standard deviation)
Violations (V) (Cronbach’s α = 0.813)
V1. I cross the street even though the pedestrian light is red.	21.6	24.1	23.8	25.3	5.2	2.7 (1.2)
V2. I cross diagonally to save time.	16.1	19.6	25.0	31.3	8.0	2.9 (1.2)
V3. I cross outside the pedestrian crossing even if there is one (crosswalk) less than 50 m away.	14.5	28.6	24.3	28.2	4.4	2.8 (1.1)
V4. I take passageways forbidden to pedestrians to save time.	33.4	27.1	16.6	18.2	4.7	2.3 (1.2)
Errors (E) (Cronbach’s α = 0.773)
E1. I cross between vehicles stopped on the roadway in traffic jams.	16.4	22.9	23.4	29.3	8.0	2.9 (1.2)
E2. I cross even if vehicles are coming because I think they will stop for me.	46.6	22.5	16.6	11.8	2.5	2.0 (1.1)
E3. I walk on cycling paths when I could walk on the sidewalk.	43.4	25.0	15.4	13.9	2.3	2.1 (1.2)
E4. I run across the street without looking because I am in a hurry.	60.9	15.5	10.0	11.3	2.3	1.8 (1.1)
Lapses (L) (Cronbach’s α = 0.910)
L1. I realize that I have crossed several streets and intersections without paying attention to traffic.	39.6	26.6	18.1	12.3	3.4	2.1 (1.2)
L2. I forget to look before crossing because I am thinking about something else.	45.9	23.4	16.2	12.0	2.5	2.0 (1.2)
L3. I cross without looking because I am talking with someone.	46.4	25.2	13.4	12.1	2.9	2.0 (1.2)
L4. I forget to look before crossing because I want to join someone on the sidewalk on the other side.	53.6	21.1	12.5	10.5	2.3	1.9 (1.1)
Aggressive behaviors (AB) (Cronbach’s α = 0.906)
AB1. I get angry with another road user (pedestrian, driver, cyclist, etc.), and I yell at him.	51.8	20.2	14.1	10.7	3.2	1.9 (1.2)
AB2. I cross very slowly to annoy a driver.	65.9	11.6	10.3	8.6	3.6	1.7 (1.2)
AB3. I get angry with another road user (pedestrian, driver, cyclist, etc.), and I make a hand gesture.	56.2	16.4	14.1	11.3	2.0	1.9 (1.1)
AB4. I have gotten angry with a driver and hit their vehicle.	73.0	7.5	8.0	8.4	3.1	1.6 (1.1)
Positive behaviors (PB) (Cronbach’s α = 0.710)
PB1. I thank a driver who stops to let me cross.	2.7	7.7	14.6	35.4	39.6	4.0 (1.0)
PB2. When I am accompanied by other pedestrians, I walk in single file on narrow sidewalks so as not to bother the pedestrians I meet.	2.7	10.2	27.7	37.9	21.6	3.6 (1.0)
PB3. I walk on the right-hand side of the sidewalk so as not to bother the pedestrians I meet.	3.1	9.1	25.9	39.6	22.3	3.7 (1.0)
PB4. I let a car go by, even if I have the right-of-way, if there is no other vehicle behind it.	4.5	15.0	28.0	37.5	15.0	3.4 (1.1)

The descriptive statistics of involvement in smartphone distraction reported by pedestrians are compiled in Table 4. The pedestrians responded to the question “How frequently do you engage in the following activities while walking?” on a scale of 1 to 5, with respect to 10 items reported in Table 4. The table shows the distribution of responses along with the mean and standard deviation of each response. The reliability coefficient of the smartphone distraction scale was found to be 0.875 indicating acceptable reliability ( 52 ). The pedestrians reported the highest frequencies for using headphones on a mobile device (mean response = 3.4) and changing music (mean response = 3.4). In contrast, the lowest frequencies were observed for talking on a mobile phone in hands-free mode (mean response = 2.7).

Table 4.

Descriptive Statistics of Involvement in Smartphone Distraction Reported by the Pedestrians

Smartphone distraction items (SD)	Responses in % (1: Never; 5: Always)					Mean (standard deviation)
Smartphone distraction items (SD)	1	2	3	4	5	Mean (standard deviation)
How frequently do you engage in the following activities while walking? (Cronbach’s α = 0.875)
SD1. Talk on mobile phone—hands free	24.3	20.0	23.0	25.0	7.7	2.7 (1.3)
SD2. Talk on mobile phone—hand held	15.5	24.3	25.4	28.6	6.2	2.9 (1.2)
SD3. Read mobile phone text message	12.1	17.0	25.0	35.7	10.2	3.1 (1.2)
SD4. Write mobile phone text message	14.8	20.9	23.6	33.0	7.7	3.0 (1.2)
SD5. Use mobile device to access Internet	12.0	13.6	27.1	33.2	14.1	3.2 (1.2)
SD6. Wear headphones on mobile device	12.7	11.1	19.8	32.9	23.6	3.4 (1.3)
SD7. Take photo or video with phone	18.1	22.1	26.4	24.5	8.9	2.8 (1.2)
SD8. Access social media	20.9	19.6	20.2	28.0	11.3	2.9 (1.3)
SD9. Change music	12.0	10.9	21.2	39.1	16.8	3.4 (1.2)
SD10. Use mobile phone for navigation	10.7	17.3	27.5	33.2	11.3	3.2 (1.1)

Structural Equation Modeling

The SEM approach is used to identify the directional relationships between the observed/measured variables and unobserved/latent variables, by processing the endogenous and exogenous variables ( 53 ). SEM involves a set of simultaneous equations to establish these relationships ( 54 ). The benefits of SEM include simultaneous estimation of all the parameter estimates in the model to identify the strength of the relationship between the variables, handling complex interrelationships involving unobserved or latent variables, and accountability of multicollinearity ( 53 ).

There are two components in an SEM model: (a) measurement model and (b) structural model ( 53 , 55 , 56 ). The measurement model is developed to define the association of the corresponding exogeneous variables with the latent variable, and the structural model is used to establish the links between the different observed and unobserved/latent variables ( 53 , 57 ). The measurement model can be written in the form of the following equations (Equation 2 is for exogenous variables and Equation 3 is for endogenous variables) ( 55 – 57 ):

X = Λ_{x} ξ + δ

(2)

Y = Λ_{y} η + ε

(3)

In the above equations, X and Y represent the respective column vectors of the observed exogenous and endogenous variables. The parameter estimates in the form of matrices are designated by Λ_x and Λ_y. The respective error terms are indicated by the column vectors δ and ε ( 55 , 57 ).

Further, the structural model can be represented in the form of the following equation (Equation 4) ( 55 ):

η = β η + γ ξ + ζ

(4)

where the variables η and ξ represent the vectors of latent endogenous and latent exogenous variables in the model. The structural coefficient matrices for latent endogenous and exogenous variables are depicted by β and γ respectively. The term ζ indicates the random variables vector ( 55 , 56 ).

In this study, path diagrams have also been developed based on the findings of the SEM models, as they are quite useful in visually representing the associations between the variables obtained from an SEM ( 58 ). Moreover, the maximum likelihood estimation method is used in this study to estimate the model parameters as suggested by past research ( 57 ). The goodness of fit of the SEMs was assessed using the comparative fit index (CFI), standardized root mean square residual (SRMR), coefficient of determination (CD), Tucker-Lewis Index (TLI), and root mean squared error of approximation (RMSEA) ( 55 – 57 ).

Findings of SEMs

In both of the SEMs, five latent variables were considered for pedestrian behavior (violations, errors, lapses, aggressive behaviors, and positive behaviors) The latent variables of pedestrian behavior were defined based on the following measured items in the questionnaire: violations (V1 to V4), errors (E1 to E4), lapses (L1 to L4), aggressive behaviors (AB1 to AB4), and positive behaviors (PB1 to PB4), as described in Table 3. The results of both SEMs are discussed in the following subsections. The overall goodness-of-fit statistics indicated a good fit for both models ( 59 ).

SEM 1: Influence of Smartphone Distraction on Road Crossing Behavior

Along with the five latent variables of pedestrian behavior mentioned above, one latent variable corresponding to smartphone distractions was considered in the first SEM. The latent variable “smartphone distractions” was defined based on the items SD1 to SD10 reported in Table 5. The standardized parameter estimates of the first SEM are presented in Table 5. Along with the estimates of variables, the table highlights the standard error, z-value, significance of the variable (p-value), and the goodness-of-fit measures at the bottom end of the table. The measurement model estimates indicate the relationships between the latent variables and their corresponding observed variables measured from the questionnaire. The structural model estimates signify the association of the “smartphone distractions” latent variable with the pedestrian behavior latent variables: “violations,”“errors,”“lapses,”“aggressive behaviors,” and “positive behaviors.” The corresponding path diagram is shown in Figure 1. In the figure, the dotted arrows indicate non-significant relationships between the variables.

Table 5.

Standardized Parameter Estimates of Structural Equation Model Involving Pedestrian Smartphone Distractions

	Variable	Estimate	Standard error	z-value	p-value
Structural model
Violations	Smartphone distractions	0.50	0.04	12.94	<0.001
Errors	Smartphone distractions	0.44	0.04	10.68	<0.001
Lapses	Smartphone distractions	0.41	0.04	10.31	<0.001
Aggressive behaviors	Smartphone distractions	0.37	0.04	8.83	<0.001
Positive behaviors	Smartphone distractions	0.04	0.05	0.80	0.424
Measurement model
V1	Violations	0.71	0.03	26.51	<0.001
V2	Violations	0.74	0.03	28.46	<0.001
V3	Violations	0.72	0.03	26.94	<0.001
V4	Violations	0.72	0.03	27.21	<0.001
E1	Errors	0.42	0.04	10.81	<0.001
E2	Errors	0.77	0.02	31.98	<0.001
E3	Errors	0.76	0.02	31.65	<0.001
E4	Errors	0.80	0.02	35.41	<0.001
L1	Lapses	0.81	0.02	46.86	<0.001
L2	Lapses	0.87	0.01	64.96	<0.001
L3	Lapses	0.86	0.01	61.86	<0.001
L4	Lapses	0.85	0.01	56.88	<0.001
AB1	Aggressive behaviors	0.81	0.02	44.14	<0.001
AB2	Aggressive behaviors	0.86	0.01	59.34	<0.001
AB3	Aggressive behaviors	0.82	0.02	46.31	<0.001
AB4	Aggressive behaviors	0.88	0.01	63.57	<0.001
PB1	Positive behaviors	0.53	0.04	13.54	<0.001
PB2	Positive behaviors	0.65	0.04	18.25	<0.001
PB3	Positive behaviors	0.76	0.03	22.02	<0.001
PB4	Positive behaviors	0.54	0.04	14.06	<0.001
SD1	Smartphone distractions	0.44	0.04	12.22	<0.001
SD2	Smartphone distractions	0.56	0.03	18.15	<0.001
SD3	Smartphone distractions	0.83	0.02	51.90	<0.001
SD4	Smartphone distractions	0.83	0.02	52.17	<0.001
SD5	Smartphone distractions	0.77	0.02	39.41	<0.001
SD6	Smartphone distractions	0.55	0.03	17.28	<0.001
SD7	Smartphone distractions	0.58	0.03	19.05	<0.001
SD8	Smartphone distractions	0.76	0.02	37.96	<0.001
SD9	Smartphone distractions	0.54	0.03	16.99	<0.001
SD10	Smartphone distractions	0.47	0.03	13.58	<0.001
Goodness-of-fit statistics
Comparative fit index		0.917
Standardized root mean square residual		0.083
Coefficient of determination		0.915
Tucker-Lewis Index		0.872
Root mean squared error of approximation		0.059

Figure 1.

Path diagram for associations between smartphone distraction and pedestrian behavior.

The measurement model results showed that all six latent variables had significant associations with their respective observed variables (factor loading > 0.4). The highest influence on “smartphone distractions” was caused by SD3 and SD4 (factor loading of both = 0.83), indicating that the smartphone distractions were best explained by reading and writing text messages (Table 5). The measured violation items had similar effects on the latent variable “violations” (factor loadings > 0.7). Similar was the case with “lapses” and “aggressive behaviors” (factor loadings > 0.8). The latent variable “errors” was best explained by E4, which indicates running across the street in a hurry without looking (factor loadings = 0.8), and least explained by E1, that is, crossing between vehicles stopped on the roadway in traffic jams (factor loadings = 0.42). Further, the “positive behaviors” latent variable had the strongest influence of PB3, that is, walking on the right-hand side of the sidewalk so as not to bother oncoming pedestrians (factor loading = 0.76) and the lowest influence of PB1, which meant thanking a driver who stops to let the pedestrian cross (factor loading = 0.53).

The structural model results revealed that smartphone distractions had the strongest influence on pedestrian road crossing violations (factor loading = 0.50), followed by errors (factor loading = 0.44), lapses (factor loading = 0.41), and aggressive behaviors (factor loading = 0.37). Interestingly, no significant effect of smartphone distractions was found on the positive road crossing behaviors of pedestrians (Figure 1).

SEM 2: Influence of Pedestrian Characteristics on Road Crossing Behavior

Table 6 summarizes the standardized parameter estimates of the second SEM model explaining the relationships of pedestrian demographics and walking characteristics with their road crossing behavior. The corresponding path analysis diagram is shown in Figure 2. In this model, along with the five latent variables of pedestrian behavior (violations, errors, lapses, aggressive behaviors, and positive behaviors), the six measured/observed variables of pedestrian characteristics obtained from the questionnaire (age, gender, crash history, walking duration, walking speed, and quality of walking environment) were included, as shown in Figure 2. The significant relationships between the variables are shown with full arrows, while the dotted arrows represent the non-significant associations.

Table 6.

Standardized Parameter Estimates of Structural Equation Model Involving Pedestrian Demographics and Walking Characteristics

	Variable	Estimate	Standar error	z-value	p-value
Structural model
Violations	Age	−0.08	0.05	−1.81	0.071
	Gender	0.12	0.05	2.79	0.005
	Crash history	0.19	0.04	4.23	<0.001
	Walking duration	−0.04	0.05	−0.83	0.405
	Quality of the walking environment	−0.12	0.05	−2.51	0.012
	Walking speed	0.13	0.05	2.75	0.006
Errors	Age	−0.04	0.04	−0.83	0.405
	Gender	0.17	0.04	3.93	<0.001
	Crash history	0.38	0.04	9.75	<0.001
	Walking duration	−0.08	0.04	−1.95	0.051
	Quality of the walking environment	0.04	0.04	0.83	0.405
	Walking speed	−0.01	0.04	−0.17	0.866
Lapses	Age	−0.04	0.04	−0.97	0.330
	Gender	0.14	0.04	3.26	0.001
	Crash history	0.30	0.04	7.48	<0.001
	Walking duration	−0.10	0.04	−2.48	0.013
	Quality of the walking environment	0.05	0.04	1.08	0.282
	Walking speed	−0.04	0.04	−0.89	0.372
Aggressive behaviors	Age	0.05	0.04	1.16	0.248
	Gender	0.15	0.04	3.68	<0.001
	Crash history	0.38	0.04	10.37	<0.001
	Walking duration	−0.06	0.04	−1.48	0.139
	Quality of the walking environment	0.03	0.04	0.67	0.505
	Walking speed	0.01	0.04	0.20	0.842
Positive behaviors	Age	0.14	0.05	2.84	0.005
	Gender	−0.13	0.05	−2.62	0.009
	Crash history	0.01	0.05	0.08	0.940
	Walking duration	−0.01	0.05	−0.27	0.784
	Quality of the walking environment	0.12	0.05	2.34	0.019
	Walking speed	0.13	0.05	2.60	0.009
Measurement model
V1	Violations	0.71	0.03	25.49	<0.001
V2	Violations	0.73	0.03	27.10	<0.001
V3	Violations	0.73	0.03	27.73	<0.001
V4	Violations	0.72	0.03	26.72	<0.001
E1	Errors	0.40	0.04	10.12	<0.001
E2	Errors	0.77	0.02	32.19	<0.001
E3	Errors	0.78	0.02	32.81	<0.001
E4	Errors	0.79	0.02	35.08	<0.001
L1	Lapses	0.81	0.02	46.49	<0.001
L2	Lapses	0.88	0.01	66.86	<0.001
L3	Lapses	0.86	0.01	61.09	<0.001
L4	Lapses	0.84	0.02	55.83	<0.001
AB1	Aggressive behaviors	0.81	0.02	44.26	<0.001
AB2	Aggressive behaviors	0.86	0.01	60.44	<0.001
AB3	Aggressive behaviors	0.82	0.02	46.09	<0.001
AB4	Aggressive behaviors	0.88	0.01	65.35	<0.001
PB1	Positive behaviors	0.53	0.04	13.78	<0.001
PB2	Positive behaviors	0.67	0.04	18.84	<0.001
PB3	Positive behaviors	0.75	0.03	22.19	<0.001
PB4	Positive behaviors	0.54	0.04	14.01	<0.001
Goodness-of-fit statistics
Comparative fit index		0.916
Standardized root mean square residual		0.072
Coefficient of determination		0.865
Tucker-Lewis Index		0.926
Root mean squared error of approximation		0.052

Figure 2.

Path diagram for associations between pedestrian demographics, walking characteristics and pedestrian behavior.

Pedestrians’ gender significantly influenced all the five pedestrian behavior variables. Male pedestrians were found to be more likely to commit violations (factor loading = 0.12), errors (factor loading = 0.17), lapses (factor loading = 0.14), and aggressive behaviors (factor loading = 0.15), whereas female pedestrians showed more positive behaviors (factor loading = −0.13). Interestingly, the influence of age was significant only on positive behaviors (factor loading = 0.14), while the effect on violations, errors, lapses, and aggressive behaviors was insignificant (Figure 2). This indicates that older/mature pedestrians are more likely to show positive behavior than young pedestrians.

The quality of the walking environment had significant associations with violations (factor loading = −0.12) and positive behaviors (factor loading = 0.12). The negative association of the walking environment quality with violations signifies that the tendency to commit violations among pedestrians decreased with better quality of the walking environment. Further, crash history had significant associations with violations (factor loading = 0.19), errors (factor loading = 0.38), lapses (factor loading = 0.30), and aggressive behaviors (factor loading = 0.38). This indicates that even though the pedestrians had prior crash experiences, they still tend to perform risky road crossing behaviors. Pedestrians who walk at high speeds were found to engage in an increased number of road crossing violations (factor loading = 0.13). However, no significant influence of walking duration was observed on any of the factors of pedestrians’ road crossing behavior (Figure 2).

Discussion

Effect of Smartphone Distraction on Pedestrian Behavior

In the present study, smartphone distraction was found to have the strongest influence on violations, followed by errors, lapses, and aggressive behaviors. This finding indicates that pedestrians engaged with their smartphones are the ones with higher road crossing violations (e.g., crossing the road even when the signal is red). As the violations are intentional behaviors, the results suggest that distracted pedestrians knowingly commit violations which is critical to their safety. The safety risk can be further enhanced when both these behaviors of using smartphones and committing road crossing violations occur together. Pedestrians’ intention has been observed to be a strong predictor of their behavior ( 60 ). In a study on Chinese pedestrians, Jiang et al. ( 23 ) found that pedestrians who are more addicted/attached to their mobile phones have higher intentions of crossing roads while distracted .

Along with the intentional violations, the other form of behaviors such as errors, lapses, and aggressiveness, were also significantly affected by technological distractions, indicating that the distraction engagement not only impairs the intentional behavior of pedestrians but also their unintentional behavior. In a study on Australian pedestrians, O’Hern et al. ( 37 ) found the highest impact of technological distraction was on violations followed by errors, which is in line with the findings of the present study. Barton et al. ( 19 ) analyzed the psychological determinants of distracted road crossing behavior of pedestrians. They found that their behavioral intentions were mainly influenced by their perceptions about control during the distracted road crossing act. As road crossing is an everyday affair for pedestrians, they may feel they can easily perform the task even while distracted by their smartphones. In some pedestrian behavior studies, pedestrian attitudes have been reported to be the major factor affecting distracted road crossing behavior ( 60 ). If pedestrians have inherent positive attitudes toward distracted road crossing behavior, they are more likely to perform the act, which may further lead to significant associations with violations, errors, lapses, and aggressive behaviors. Another interesting finding of this study was that the frequency of smartphone distraction did not significantly influence the positive behaviors among pedestrians (e.g., thanking a driver who stopped their vehicle to let the pedestrian pass). However, the pedestrians reported the the highest frequencies of performing positive behaviors compared with the other behavioral aspects, which showcases the civilized nature of Indian pedestrians, similar to past research conducted on American, French, Iranian, and Serbian pedestrians ( 17 , 30 , 36 , 39 ).

Effect of Pedestrian Characteristics on Pedestrian Behavior

Age was found to be a significant influencing factor in the positive behaviors of pedestrians, indicating that older/mature pedestrians are more likely to show positive behavior than young pedestrians ( 60 ). In a study on pedestrians from LMICs, McIlroy et al. ( 29 ) reported that older people perform better than younger people when subjected to road crossing tasks. Deb et al. ( 30 ) revealed age as a significant factor influencing all five aspects of pedestrian behavior (violations, errors, lapses, aggressive and positive behaviors) among American pedestrians. With reference to Greek pedestrians, Papadimitriou et al. ( 4 ) found that younger pedestrians committed higher rates of errors and violations. However, in the present study, no significant influence of age was found on errors and violations of pedestrians. Interestingly, male pedestrians were more likely to commit violations, errors, lapses, and aggressive behaviors, whereas female pedestrians showed more positive behaviors. This shows that males tend to perform risky road crossing behaviors more than females ( 39 , 61 ). An investigation on American pedestrians by Deb et al. ( 30 ) also indicated a similar finding with respect to gender effects on pedestrian behavior. In Serbia, Antić et al. ( 35 ) found that young pedestrians committed the highest number of violations, whereas pedestrians who walk for short distances commit the highest number of lapses. Moreover, they reported female and old-aged pedestrians to show a higher rate of positive behaviors compared with other pedestrians.

Another observation of the present study was that pedestrians with fast walking speeds were more likely to engage in a higher frequency of violations. However, observational research on pedestrian behavior has demonstrated that pedestrians with slower walking speeds are more likely to end up in road violations while crossing ( 62 ). Moreover, pedestrians with a crash history were more likely to commit violations and show aggressive behaviors while crossing the road than those with no prior crash experience. This finding was in consensus with the observations of a previous study by McIlroy et al. ( 29 ), who reported that pedestrians who indulged in risky road crossing behaviors had a higher likelihood of prior crash history. However, the impact of walking duration was insignificant for all five factors of pedestrians’ road crossing behavior. Moreover, a better walking environment led to a lower frequency of pedestrian violations and better positive behaviors, which complements the past research stating that pedestrian crashes were reduced with better urban planning policies and improved built environment features ( 49 ).

Practical Implications

The study findings can aid policymakers in designing behavior-based interventions to minimize road traffic crashes involving pedestrians. Figure 3 illustrates a flowchart depicting the suggested approach to assist in developing pedestrian safety countermeasures and designing ITS-based applications to minimize the influence of smartphone distraction on road safety. In this context, the first step is to investigate the frequency of smartphone distraction among pedestrians, covering all the various types of audio/visual/manual distraction (e.g., talking, texting, listening to music, etc.) along with their demographic characteristics and walking attributes. The second step is to assess their road crossing behavior with the help of behavioral characteristics (violations, errors, lapses, aggressive and positive behaviors). The third step can be categorized into two parts: (a) quantification of the impact of frequency of smartphone distraction on pedestrian behavior, and (b) quantification of the influence of demographic and walking attributes on their behavior. In the fourth step, pedestrians for whom the impact of distraction on behavior is found to be high can be termed “highly distracted” pedestrians, and those who are critical with regard to their demographic and walking attributes can be called “highly vulnerable” pedestrians. Both these categories of pedestrians are more likely to be involved in a road crash compared with other pedestrians. After identifying the at-risk category of pedestrians, their identification characteristics can be used as inputs for ITS applications utilized in road safety mitigation and prevention. Accordingly, pedestrians can be alerted of their crash risk through advanced pedestrian warning systems such as bluetooth beacons ( 63 ), illuminated in-ground lights ( 64 ), and so forth. Simultaneously, drivers can be alerted through collision avoidance systems.

Figure 3.

Flowchart depicting the suggested approach to aid in developing road safety countermeasures.

Limitations and Future Research

There were a few limitations observed in the present study. As the mode of conducting this study was survey-based, the responses of pedestrians about their behavior may not be exactly the same as their behavior in the field since a lot of situational traffic and environmental factors may influence their behavior. As a result of the online data collection and English as a language of the survey, people not well versed with the technology and language could not participate in the study. In this study, only adult pedestrians (18 years and above) were considered, which would have left out the child pedestrians, thereby influencing the generalizability of the study findings. Therefore, future studies should inculcate more robust sampling approaches and consider a more comprehensive demographic range of participants. Another limitation is that the area type (rural or urban) and the crosswalk type (signalized or unsignalized or midblock intersections) which the pedestrians frequently use were not enquired about in the study, which may influence their behavioral responses. Further, pedestrians’ road crossing behavior is also influenced by the type of vehicle approaching them at crosswalks. Future studies can investigate the interaction between distraction engagement and vehicle type in influencing pedestrian safety.

Conclusions

Road traffic crashes involving pedestrians are more common in developing countries than developed nations. With the increasing use of technology, smartphone distraction has emerged as a significant distracting activity observed among pedestrians ( 65 ). In recent years, research on pedestrian behavior has become quite prevalent in developed countries; however, it is scarce in LMICs. The present study is the first attempt to investigate the effect of smartphone distraction on pedestrians’ road crossing behavior in India. The study confirms the efficacy of PBQ in investigating safety-related behavior among Indian pedestrians. The strongest influence of smartphone distraction was found to be on violations, followed by errors, lapses, and aggressive behaviors. Male pedestrians were more likely to commit violations, errors, lapses, and aggressive behaviors, whereas female pedestrians showed more positive behaviors. Older pedestrians showed more positive behavior than young pedestrians. Pedestrians who walk at high speeds tend to engage in a high number of violations. Pedestrians with a crash history were more likely to commit violations and show aggressive behaviors than those with no prior crash experience. These study findings can be helpful in understanding the smartphone distraction habits of pedestrians while making road crossing decisions and in the development of pedestrian safety countermeasures.

Footnotes

Author Contributions

The authors confirm their contribution to the paper as follows: study conception and design: A. K. Yadav, N. M. Pawar, N. R. Velaga; data collection: A. K. Yadav, N. M. Pawar; analysis and interpretation of results: A. K. Yadav; draft manuscript preparation: A. K. Yadav, N. M. Pawar, N. R. Velaga. All authors reviewed the results and approved the final version of the manuscript.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iDs

Ankit Kumar Yadav

Nishant Mukund Pawar

Nagendra R. Velaga

Data Accessibility Statement

Data will be available on request.

References

Al Shammas

Escobar

Comfort and Time-Based Walkability Index Design: A GIS-Based Proposal. International Journal of Environmental Research and Public Health, Vol. 16, No. 16, 2019, p. 2850.

Kelly

Williamson

Niven

A. G.

Hunter

Mutrie

Richards

Walking on Sunshine: Scoping Review of the Evidence for Walking and Mental Health. British Journal of Sports Medicine, Vol. 52, No. 12, 2018, pp. 800–806.

Perchoux

Brondeel

Wasfi

Klein

Caruso

Vallée

Klein

, et al. Walking, Trip Purpose, and Exposure to Multiple Environments: A Case Study of Older Adults in Luxembourg. Journal of Transport & Health, Vol. 13, 2019, pp. 170–184.

Papadimitriou

Lassarre

Yannis

Introducing Human Factors in Pedestrian Crossing Behaviour Models. Transportation Research Part F: Traffic Psychology and Behaviour, Vol. 36, 2016, pp. 69–82.

Yannis

Nikolaou

Laiou

Stürmer

Y. A.

Buttler

Jankowska-Karpa

Vulnerable Road Users: Cross-Cultural Perspectives on Performance and Attitudes. IATSS Research, Vol. 44, No. 3, 2020, pp. 220–229.

World Health Organization. Global Status Report on Road Safety 2018: Summary. No. WHO/NMH/NVI/18.20. World Health Organization, Geneva, 2018.

Koh

P. P.

Wong

Y. D.

Chandrasekar

Safety Evaluation of Pedestrian Behavior and Violations at Signalised Pedestrian Crossings. Safety Science, Vol. 70, 2014, pp. 143–152.

Brosseau

Zangenehpour

Saunier

Miranda-Moreno

The Impact of Waiting Time and Other Factors on Dangerous Pedestrian Crossings and Violations at Signalized Intersections: A Case Study in Montreal. Transportation Research Part F: Traffic Psychology and Behavior, Vol. 21, 2013, pp. 159–172.

Alver

Onelcin

Cicekli

Abdel-Aty

Evaluation of Pedestrian Critical Gap and Crossing Speed at Midblock Crossing Using Image Processing. Accident Analysis & Prevention, Vol. 156, 2021, p. 106127.

10.

Mukherjee

Mitra

A Comprehensive Study on Factors Influencing Pedestrian Signal Violation Behavior: Experience from Kolkata City, India. Safety Science, Vol. 124, 2020, p. 104610.

11.

Pawar

D. S.

Yadav

A. K.

Modelling the Pedestrian Dilemma Zone at Uncontrolled Midblock Sections. Journal of Safety Research, Vol. 80, 2022, pp. 87–96.

12.

Yadav

A. K.

Velaga

N. R.

A Systematic Review of Observational Studies Investigating the Influence of Mobile Phone Distraction on Road Crossing Behaviour of Pedestrians. Transportation Research Part F: Traffic Psychology and Behaviour, Vol. 91, 2022, pp. 236–259.

13.

Scopatz

R. A.

Zhou

Effect of Electronic Device Use on Pedestrian Safety: A Literature Review. National Highway Traffic Safety Administration, Washington, D.C., 2016.

14.

Campisi

Otković

I. I.

Šurdonja

Deluka-Tibljaš

Impact of Social and Technological Distraction on Pedestrian Crossing Behaviour: A Case Study in Enna, Sicily. Transportation Research Procedia, Vol. 60, 2022, pp. 100–107.

15.

Simmons

S. M.

Caird

J. K.

Sterzer

Hagel

B. E.

Plight of the Distracted Pedestrian: A Research Synthesis and Meta-Analysis of Mobile Phone Use on Crossing Behaviour. Injury Prevention, Vol. 26, No. 2, 2020, pp. 170–176.

16.

Sobhani

Farooq

Impact of Smartphone Distraction on Pedestrians’ Crossing Behaviour: An Application of Head-Mounted Immersive Virtual Reality. Transportation Research Part F: Traffic Psychology and Behaviour, Vol. 58, 2018, pp. 228–241.

17.

Pešić

Antić

Glavić

Milenković

The Effects of Mobile Phone Use on Pedestrian Crossing Behaviour at Unsignalized Intersections - Models for Predicting Unsafe Pedestrians Behaviour. Safety Science, Vol. 82, 2016, pp. 1–8.

18.

Nasar

J. L.

Troyer

Pedestrian Injuries Due to Mobile Phone Use in Public Places. Accident Analysis & Prevention, Vol. 57, 2013, pp. 91–95.

19.

Barton

B. K.

Kologi

S. M.

Siron

Distracted Pedestrians in Crosswalks: An Application of the Theory of Planned Behavior. Transportation Research Part F: Traffic Psychology and Behaviour, Vol. 37, 2016, pp. 129–137.

20.

Liu

Alsaleh

Sayed

Modeling the Influence of Mobile Phone Use Distraction on Pedestrian Reaction Times to Green Signals: A Multilevel Mixed-Effects Parametric Survival Model. Transportation Research Part F: Traffic Psychology and Behaviour, Vol. 81, 2021, pp. 115–129.

21.

Park

Kim

Development and Validation of a Novel Instrument to Measure Pedestrians’ Smartphone Use: The Smombie Scale. Transportation Research Part F: Traffic Psychology and Behaviour, Vol. 82, 2021, pp. 440–449.

22.

Young

K. L.

Salmon

P. M.

Examining the Relationship Between Driver Distraction and Driving Errors: A Discussion of Theory, Studies and Methods. Safety Science, Vol. 50, No. 2, 2012, pp. 165–174.

23.

Jiang

Ling

Feng

Wang

Guo

Psychological Predictors of Mobile Phone Use While Crossing the Street Among College Students: An Application of the Theory of Planned Behavior. Traffic Injury Prevention, Vol. 18, No. 2, 2017, pp. 118–123.

24.

Lennon

Oviedo-Trespalacios

Matthews

Pedestrian Self-Reported Use of Smart Phones: Positive Attitudes and High Exposure Influence Intentions to Cross the Road While Distracted. Accident Analysis & Prevention, Vol. 98, 2017, pp. 338–347.

25.

Lin

M. I. B.

Huang

Y. P.

The Impact of Walking While Using a Smartphone on Pedestrians’ Awareness of Roadside Events. Accident Analysis & Prevention, Vol. 101, 2017, pp. 87–96.

26.

Yadav

A. K.

Choudhary

S. S.

Pawar

N. M.

Velaga

N. R.

Investigating the Role of Beliefs in Influencing the Hand-Held and Hands-Free Mobile Phone Use Among Pedestrians in India. International Journal of Injury Control and Safety Promotion, Vol. 30, No. 1, 2023, pp. 79–90.

27.

Wang

Zhang

, et al. Driver distraction detection based on vehicle dynamics using naturalistic driving data. Transportation Research Part C: Emerging Technologies, Vol. 136, 2022, p. 103561.

28.

Ponte

Edwards

S. A.

Wundersitz

The Prevalence of In-Vehicle Driver Distraction in Moving Traffic. Transportation Research Part F: Traffic Psychology and Behaviour, Vol. 83, 2021, pp. 33–41.

29.

McIlroy

R. C.

Plant

K. L.

Jikyong

Nam

V. H.

Bunyasi

Kokwaro

G. O.

Hoque

M. S.

Preston

J. M.

Stanton

N. A.

. Vulnerable Road Users in Low-, Middle-, and High-Income Countries: Validation of a Pedestrian Behaviour Questionnaire. Accident Analysis & Prevention, Vol. 131, 2019, pp. 80–94.

30.

Deb

Strawderman

DuBien

Smith

Carruth

D. W.

Garrison

T. M.

Evaluating Pedestrian Behavior at Crosswalks: Validation of a Pedestrian Behavior Questionnaire for the U.S. Population. Accident Analysis & Prevention, Vol. 106, 2017, pp. 191–201.

31.

Reason

Manstead

Stradling

Baxter

Campbell

Errors and Violations on the Roads: A Real Distinction?

Ergonomics, Vol. 33, No. 10–11, 1990, pp. 1315–1332.

32.

Parker

Reason

J. T.

Manstead

A. S.

Stradling

S. G.

Driving Errors, Driving Violations and Accident Involvement. Ergonomics, Vol. 38, No. 5, 1995, pp. 1036–1048.

33.

Lawton

Parker

Manstead

A. S.

Stradling

S. G.

The Role of Affect in Predicting Social Behaviors: The Case of Road Traffic Violations. Journal of Applied Social Psychology, Vol. 27, No. 14, 1997, pp. 1258–1276.

34.

Özkan

Lajunen

What Causes the Differences in Driving Between Young Men and Women? The Effects of Gender Roles and Sex on Young Drivers’ Driving Behaviour and Self-Assessment of Skills. Transportation Research Part F: Traffic Psychology and Behaviour, Vol. 9, No. 4, 2006, pp. 269–277.

35.

Antić

Pešić

Milutinović

Maslać

Pedestrian Behaviours: Validation of the Serbian Version of the Pedestrian Behaviour Scale. Transportation Research Part F: Traffic Psychology and Behaviour, Vol. 41, 2016, pp. 170–178.

36.

Esmaili

Aghabayk

Parishad

Stephens

A. N.

Investigating the Interaction Between Pedestrian Behaviors and Crashes Through Validation of a Pedestrian Behavior Questionnaire (PBQ). Accident Analysis & Prevention, Vol. 153, 2021, p. 106050.

37.

O’Hern

Stephens

A. N.

Estgfaeller

Moore

Koppel

Self-Reported Pedestrian Behaviour in Australia. Transportation Research Part F: Traffic Psychology and Behaviour, Vol. 75, 2020, pp. 134–144.

38.

Useche

S. A.

Alonso

Montoro

Validation of the Walking Behavior Questionnaire (WBQ): A Tool for Measuring Risky and Safe Walking Under a Behavioral Perspective. Journal of Transport and Health, Vol. 18, 2020, p. 100899.

39.

Granié

M. A.

Pannetier

Guého

Developing a Self-Reporting Method to Measure Pedestrian Behaviors at All Ages. Accident Analysis & Prevention, Vol. 50, 2013, pp. 830–839.

40.

Nordfjærn

Şimşekoğlu

Ö.

The Role of Cultural Factors and Attitudes for Pedestrian Behaviour in an Urban Turkish Sample. Transportation Research Part F: Traffic Psychology and Behaviour, Vol. 21, 2013, pp. 181–193.

41.

Rodríguez

A. H.

Avilés-Jurado

F. X.

Díaz

Schuetz

Trefler

S. I.

Solé-Violán

Cordero

, et al. Procalcitonin (PCT) Levels for Ruling-Out Bacterial Coinfection in ICU Patients with Influenza: A CHAID Decision-Tree Analysis. Journal of Infection, Vol. 72, No. 2, 2016, pp. 143–151.

42.

Elliott

M. A.

Baughan

C. J.

Developing a Self-Report Method for Investigating Adolescent Road User Behaviour. Transportation Research Part F: Traffic Psychology and Behaviour, Vol. 7, No. 6, 2004, pp. 373–393.

43.

Díaz

E. M.

Theory of Planned Behavior and Pedestrians’ Intentions to Violate Traffic Regulations. Transportation Research Part F: Traffic Psychology and Behaviour, Vol. 5, No. 3, 2002, pp. 169–175.

44.

Elliott

M. A.

Armitage

C. J.

Baughan

C. J.

Drivers’ Compliance with Speed Limits: An Application of the Theory of Planned Behavior. Journal of Applied Psychology, Vol. 88, No. 5, 2003, pp. 964–972.

45.

Zhang

Zhao

Zhang

The Effect of Cognitive Errors, Mindfulness and Personality Traits on Pedestrian Behavior in a Chinese Sample. Transportation Research Part F: Traffic Psychology and Behaviour, Vol. 41, 2016, pp. 29–37.

46.

Young

K. L.

Koppel

Stephens

A. N.

Osborne

Chambers

Hassed

Mindfulness Predicts Driver Engagement in Distracting Activities. Mindfulness, Vol. 10, No. 5, 2019, pp. 913–922.

47.

Holland

Hill

Gender Differences in Factors Predicting Unsafe Crossing Decisions in Adult Pedestrians Across the Lifespan: A Simulation Study. Accident Analysis & Prevention, Vol. 42, No. 4, 2010, pp. 1097–1106.

48.

Lee

Abdel-Aty

Huang

Cai

Transportation Safety Planning Approach for Pedestrians: An Integrated Framework of Modeling Walking Duration and Pedestrian Fatalities. Transportation Research Record: Journal of the Transportation Research Board, 2019. 2673: 898–906.

49.

Congiu

Sotgiu

Castiglia

Azara

Piana

Saderi

Dettori

Built Environment Features and Pedestrian Accidents: An Italian Retrospective Study. Sustainability (Switzerland), Vol. 11, No. 4, 2019, pp. 1–14.

50.

Aghabayk

Esmailpour

Jafari

Shiwakoti

Observational-Based Study to Explore Pedestrian Crossing Behaviors at Signalized and Unsignalized Crosswalks. Accident Analysis & Prevention, Vol. 151, 2021, p. 105990.

51.

Cochran

W. G.

Sampling Techniques, 3rd ed. John Wiley & Sons, New York, NY, 1977.

52.

Nunnally

Psychometric Methods. McGraw-Hill, New York, NY, 1978.

53.

Hassan

H. M.

Abdel-Aty

M. A.

Analysis of Drivers’ Behavior Under Reduced Visibility Conditions Using a Structural Equation Modeling Approach. Transportation Research Part F: Traffic Psychology and Behavior, Vol. 14, No. 6, 2011, pp. 614–625.

54.

Lee

J. Y.

Chung

J. H.

Son

Analysis of Traffic Accident Size for Korean Highway Using Structural Equation Models. Accident Analysis & Prevention, Vol. 40, No. 6, 2008, pp. 1955–1963.

55.

Zhao

Chen

An Analysis of the Relationship Between Driver Characteristics and Driving Safety Using Structural Equation Models. Transportation Research Part F: Traffic Psychology and Behavior, Vol. 62, 2019, pp. 529–545.

56.

Yadav

A. K.

Velaga

N. R.

An Investigation on the Risk Factors Associated with Driving Errors Under the Influence of Alcohol Using Structural Equation Modeling. Traffic Injury Prevention, Vol. 21, No. 4, 2020, pp. 288–294.

57.

Papantoniou

Structural Equation Model Analysis for the Evaluation of Overall Driving Performance: A Driving Simulator Study Focusing on Driver Distraction. Traffic Injury Prevention, Vol. 19, No. 3, 2018, pp. 317–325.

58.

Wright

The Method of Path Coefficients. The Annals of Mathematical Statistics, Vol. 5, No. 3, 1934, pp. 161–215.

59.

Bentler

P. M.

Bonett

D. G.

Significance Tests and Goodness of Fit in the Analysis of Covariance Structures. Psychological Bulletin, Vol. 88, No. 3, 1980, pp. 588–606.

60.

Hou

Cheng

Xiao

Wang

Distracted Behavior of Pedestrians While Crossing Street: A Case Study in China. International Journal of Environmental Research and Public Health, Vol. 18, No. 1, 2021, pp. 1–19.

61.

Dinh

D. D.

N. H.

McIlroy

R. C.

Plant

K. A.

Stanton

N. A.

Examining the Roles of Multidimensional Fatalism on Traffic Safety Attitudes and Pedestrian Behaviour. Safety Science, Vol. 124, 2020, p. 104587.

62.

Russo

B. J.

James

Aguilar

C. Y.

Smaglik

E. J.

Pedestrian Behavior at Signalized Intersection Crosswalks: Observational Study of Factors Associated with Distracted Walking, Pedestrian Violations, and Walking Speed. Transportation Research Record: Journal of the Transportation Research Board, 2018. 2672: 1–12.

63.

Hasan

Hoque

M. A.

Karim

Griffin

Schwebel

D. C.

Hasan

, Someone to Watch Over You: Using Bluetooth Beacons for Alerting Distracted Pedestrians. IEEE Internet of Things, Vol. 9, No. 22, 2022, pp. 23017–23030.

64.

Larue

G. S.

Watling

C. N.

Black

Wood

J. M.

Improving the Safety of Distracted Pedestrians with In-Ground Flashing Lights. A Railway Crossing Field Study. Journal of Safety Research, Vol. 77, 2021, pp. 170–181.

65.

Rahmillah

F. I.

Tariq

King

Oviedo-Trespalacios

Is Distraction on the Road Associated with Maladaptive Mobile Phone Use? A Systematic Review. Accident Analysis & Prevention, Vol. 181, 2023, p. 106900.