Understanding Distraction Types: Comparative Analysis of In-Vehicle and External Driver Distractions in Pedestrian Fatal Crashes

Abstract

Distracted driving remains a significant public safety concern, contributing to numerous severe injuries and fatalities in the world. This study analyzes 8 years of Fatality Analysis Reporting System data (2016 to 2023) from the United States using association rule mining to identify patterns associated with various types of driver distractions, specifically distinguishing between external and in-vehicle distractions. The findings indicated that in-vehicle distractions were predominantly associated with young drivers, drug and alcohol use, daylight conditions, arterial roads, high-speed limits, sport utility vehicles, and vehicles of recent model years (2019 to 2024). In contrast, external distractions were more commonly linked to middle-aged and senior drivers, local roads, lower speed limits, dark and unlit conditions, trucks, vans, buses, and vehicles from earlier model years (1980 to 1995). The study recommends strengthening primary enforcement of distracted-driving laws. Since several distracted-driving laws target in-vehicle distractions such as the use of handheld electronic devices, it is imperative to promote advanced driver-assistance technologies in newer vehicles that can mitigate external distractions in addition to in-vehicle distractions. Additional measures include improving signage, fencing, roadside assistance, and visibility in low-light areas, alongside implementing age-targeted educational campaigns to address distraction risks across different driver groups.

Keywords

driver distraction pedestrians association fatality

Introduction

Driver distraction is a significant contributor to road crashes globally. It occurs when a driver’s attention is diverted from the primary task of driving to secondary activities, such as conversing with passengers, using mobile devices, or visually fixating on external events like roadside activities and objects. Qin et al. found that in 2017, 476 pedestrian fatalities and 30,000 injuries were estimated to be from cellphone-related distracted driving alone ( 1 ). According to the National Highway Traffic Safety Administration (NHTSA), more than 3,000 lives were lost to distracted driving in the United States in 2023. Of these, 492 fatal crashes involved distracted drivers striking pedestrians ( 2 ). Analysis of data from the U.S. Fatality Analysis Reporting System (FARS) indicates that this trend has remained consistent over the past 5 years, with approximately 500 pedestrian fatalities annually attributed to driver distraction.

Driver distractions are commonly classified as visual, manual, or cognitive ( 2 ). Visual distractions involve looking away from the road, manual distractions involve removing hands from the steering wheel; and cognitive distractions occur when attention shifts away from driving despite having eyes on the road and hands on the wheel. Driver distractions can originate from either internal sources (such as mobile phone use, eating, or adjusting vehicle controls) or external sources (including roadside billboards, pedestrians, or events occurring outside the vehicle).

Previous research found that distraction types can be influenced by sociodemographic factors such as age and gender. For example, young drivers, especially females, are more likely to be distracted by in-vehicle technology ( 3 , 4 ). The results indicated that male pedestrians aged 25 to 64 were struck by a distracted driver in urban areas. This study also found that pedestrian fatalities per 10-billion vehicle miles traveled increased from 116.1 (2005) to 168.6 (2010) as a result of distracted driving ( 3 ).

Although extensive research has examined distraction broadly ( 4 , 5 ), few studies focus on the specific roles of different types of distraction in pedestrian crashes. Given the growing concerns about pedestrian safety, it is essential to investigate how contextual, behavioral, and environmental factors influence different types of distraction and their outcomes. Examining the different types of distraction is crucial to understanding how distraction causes specific types of crashes and for designing effective interventions. For example, addressing external distracted drivers continues to be challenging because most distracted-driving laws primarily target observable, device-based behaviors, such as handheld phone use, rather than cognitive or visually driven distractions originating outside the vehicle.

External distractions, such as roadside advertising, digital billboards, crashes on the shoulder, or pedestrian activity, are difficult to detect and therefore regulate because they involve elements of the roadway environment that are not directly under the driver’s control ( 2 ). Furthermore, unlike in-vehicle device use, external distractions are harder to detect and enforce because they leave no physical evidence and do not involve clear, illegal actions by the driver ( 6 ). Existing statutes therefore struggle to address these forms of distraction, even though external visual or cognitive demands can significantly impair driver performance.

This study addresses the need to analyze the contextual and behavioral factors associated with distracted driving that result in pedestrian fatalities. Therefore, the main objective of this study is to identify and examine the factors contributing to different types of driver distraction that result in pedestrian fatalities. The goal is to highlight the significance of understanding the types of distractions and, having done so, suggest potential interventions that address them.

Literature Review

Distracted driving is a diverse safety concern affected by driver behavior, vehicle features, and roadway environments. This section reviews key literature on in-vehicle, external, and general distraction, and on pedestrian safety to identify current insights and remaining gaps as summarized below.

Studies on In-Vehicle Distraction

Research on in-vehicle distractions has used crash data, observational methods, and simulator experiments to understand how cellphone use and related behaviors affect driving performance and safety. According to studies, factors such as driver characteristics, roadway conditions, and specific phone tasks consistently emerge as significant contributors to distraction-related risk. A previous study examined in-vehicle distraction sources using a Bayesian multinomial logit model. The study utilized 5,078 distracted-driving crash records from Iowa between 2015 and 2018. The results showed that older drivers, alcohol use, and higher speed limits increased the likelihood of cellphone-related distraction crashes ( 7 ). Another study by Hasan et al. analyzed crash data from 2015 to 2019 involving cellphone use in New Jersey using mixed logit models. The results showed that higher speed limits and a larger total number of vehicles involved increased crash severity ( 8 ).

In another study, researchers investigated cellphone-related distracted driving using observational data from 3,727 drivers and applied association rule mining (ARM). Their findings showed that passenger car drivers and female drivers traveling on continuous road segments were more frequently associated with phone conversations ( 9 ). Another study used a driving simulator to assess the impact of phone use on driving performance among young and professional drivers. The findings showed that texting impaired lane control for both drivers, with the effect being more pronounced among younger drivers ( 10 ).

Studies on External Distraction

External distractions have emerged as a significant concern in understanding driver behavior, as roadside elements increasingly compete for drivers’ visual and cognitive attention. Recent studies highlight the growing influence of digital billboards and other environmental stimuli on driving performance, underscoring the need to examine how such factors contribute to distraction-related risk. For example, Brome et al. investigated the effect of digital billboard advertisements on driver performance and attention using driving simulators. Their results indicated that digital billboards significantly increased drivers’ cognitive load ( 11 ). Another study conducted a naturalistic driving study in Iran and applied structural equation modeling to assess human, road, and environmental factors influenced by digital billboards. Their findings showed that younger male drivers, nighttime or rainy conditions, and locations near intersections were associated with a higher likelihood of driver distraction ( 12 ).

Studies on General Distraction

Recent research has consistently shown that driver distraction is a multifaceted safety challenge influenced by driver characteristics, roadway environments, and the nature of the distracting stimuli. Studies using diverse methods, including crash data modeling, naturalistic driving observations, roadside surveys, and driving simulators, demonstrate that both in-vehicle and external distractions significantly shape crash risk and driving performance. Collectively, the research highlights how distraction patterns vary across age groups, vehicle types, and road contexts, providing a foundation for understanding the complex mechanisms through which distraction contributes to unsafe driving outcomes.

A previous study analyzed distracted-driving injury severities across different vehicle types using random-parameter multinomial logit models. The study found that severe injuries in passenger cars and SUVs were more commonly associated with female drivers and urban roads, whereas in pickup trucks and minivans severe injuries were linked to male drivers and rural environments ( 13 ). Another study by Guo et al. quantified crash risk associated with different distractions across driver age groups. The study used mixed-effect logistic regression on data from naturalistic driving. The results depicted visual–manual distractions affecting drivers of all ages. Cellphone use was highest among young adults, whereas talking to passengers and looking outside was common among teens ( 14 ).

The prevalence of driver distraction was estimated in a roadside observational study in Alabama. The observation showed that texting was more frequent on high-speed roads and while stopped at intersections. External distractions occurred more on local roads, whereas texting was more common on arterials or collector roads ( 15 ). Horberry et al. used a driving simulator to examine how in-vehicle tasks, road complexity, and driver age affect driving performance. They found that radio use impaired driving performance, and that environments with heavier traffic, buildings, and visual stimuli led to reduced driving speeds, particularly among older drivers ( 16 ).

Pedestrian-Related Studies

Recent research on pedestrian safety has explored a range of factors contributing to crash occurrence, severity, and mitigation across diverse roadway environments. Studies have examined pedestrian crash patterns, behavioral risk factors, roadway design influences, and stakeholder perceptions, collectively highlighting the complex interactions between environmental conditions, driver behavior, and infrastructure characteristics. However, these studies have not specifically addressed pedestrian crashes related to distracted driving, especially from the perspective of distraction types. A previous study by Das et al. evaluated vehicle–pedestrian crash patterns in urban areas using 8 years of Louisiana crash data (2004 to 2011) with the aid of ARM. The findings indicated that crashes occurring in dark conditions were likely to be fatal. Middle-aged male pedestrians were associated with impaired conditions ( 17 ).

Another study identified risk factors for fatal pedestrian crashes from 2009 to 2019 in West Virginia. The study employed multivariable logistic regression and spatial analysis. The results showed that most fatalities occurred at night, on dry roads, and involved alcohol use. Male crashes were more spatially concentrated than female crashes ( 18 ). Other studies have evaluated the effect of roadway features on pedestrian safety; for instance, Novat et al. examined the influence of slip lane designs on pedestrian safety in light of the increase of autonomous vehicles. The study developed a Vissim simulation model using vehicle trajectory data collected in Michigan. The results showed that yield-controlled slip lanes produced lower vehicle speeds and improved pedestrian safety compared with a free-flow design ( 19 ). In a survey of 206 planning professionals, 156 police officers, and 788 pedestrians in New Jersey, United States, distracted driving was viewed as a very serious issue by 98% of professionals, 99% of police officers, and 89% of pedestrians ( 20 ).

Study Contribution to the State of the Art

Although distracted driving has been widely studied, limited research has specifically investigated how distinct types of driver distraction contribute to fatal pedestrian crashes. Existing literature often treats distraction as a general factor, with minimal differentiation between in-vehicle and external sources. This study advances the state of knowledge by disaggregating distraction sources and applying ARM to national FARS data (2016 to 2023) to uncover unique contextual and behavioral patterns associated with each type. It highlights that reducing crash risks associated with external distractions will require strategies beyond traditional enforcement-based approaches, including roadway design, environmental controls, advancement in technology, and emphasizing public education. The study also highlights the need for policy frameworks that more fully account for the cognitive and visual demands imposed by the driving environment, not just in-vehicle behaviors.

Data

The study used crash data from 2016 to 2023, stored and collected by FARS. FARS is a database maintained by NHTSA that contains information on all fatal motor vehicle crashes within the United States. The study obtained the crash data and filtered it to identify pedestrian crashes associated with driver distractions.

Data Preparation

The data preparation begins at the crash level, where each crash in the “accident” file is identified using the unique identifier “ST_CASE.” This identifier is then used to connect each crash to the vehicle level, which provides details on every vehicle involved from the “vehicle” file and the distraction status of the driver from the “distract” file using both “ST_CASE” and “VEH_NO.” It is then linked to the person level, which contains information on the drivers, passengers, and nonmotorists involved in the crash identified in the “person” file and the “pbtype” file (i.e., data file on pedestrians and bicyclists,) using “ST_CASE,”“PER_NO,” and “VEH_NO” as presented in Figure 1. The categories of distractions from the distract file were grouped into in-vehicle and external distractions as described in Table 1.

Table 1.

Attribute Elements in FARS Analytical User’s Manual 1975 to 2021

Attribute name	Code
In-vehicle distraction
By other occupant(s)	3
By a moving object in vehicle	4
While talking or listening to mobile phone	5
While manipulating mobile phone	6
While adjusting audio or climate controls	7
While using other component/controls integral to vehicle	9
While using or reaching for device/object brought into vehicle	10
Eating or drinking	13
Smoking-related	14
Other mobile phone–related	15
External distraction
Distracted by outside person, object, or event	12

Note: FARS = Fatality Analysis Reporting System.

The final crash dataset contained crashes involving one distracted driver (by either in-vehicle or external distraction) who collided with one pedestrian. Drivers with multiple distraction sources, and crashes involving more than one driver or pedestrian, were excluded to ensure clear analysis and to simplify interpretation.

After eliminating records with missing information, a total of 1,286 crash cases remained and were used for further analysis, in which 804 crashes were related to in-vehicle distractions, and 482 crashes were related to external distractions. The flow chart detailing the analytical process is shown in Figure 2.

Descriptive Statistics of the Data

To analyze the contributing factors for different distraction types, 23 variables (including the distraction type) were selected based on information extracted from crash reports, findings from the literature, and established engineering practices. The variables were grouped into driver characteristics, pedestrian characteristics, roadway characteristics, environmental characteristics, and vehicle characteristics. The data included the “unknown” variable, which represents missing or incomplete information in certain crash records and was maintained to ensure consistency in data analysis and interpretation.

Table 2 presents the descriptive statistics of the 23 variables by comparing the two distraction types. It was observed that the percentage for driver’s age category of 26 to 45 (young adult) in in-vehicle distracted crashes (49.63%) was larger than with external distracted crashes (38.80%). However, for middle-aged drivers (aged 46 to 64), higher proportions were observed in external distracted crashes (30.08%) compared with in-vehicle distracted crashes (16.42%). Female drivers exhibited a higher proportion of in-vehicle distractions (32.71%) compared with external distractions (26.56%). A higher percentage of in-vehicle distracted driver crashes (71.39%) occurred on arterial roads compared with externally distracted driver crashes (64.32%).

Table 2.

Descriptive Statistics of the Dataset

Variables	In-vehicle distraction		External distraction
Variables	Count	Percent	Count	Percent
Driver characteristics
Driver age
0–25 (young)	236	29.35	97	20.12
26–45 (young adult)	399	49.63	187	38.80
46–64 (middle aged)	132	16.42	145	30.08
65+ (senior)	29	3.61	49	10.17
Unknown	8	1.00	4	0.83
Driver gender
Male	532	66.17	349	72.41
Female	263	32.71	128	26.56
Unknown	9	1.12	5	1.04
Driver alcohol use
No	732	91.04	456	94.61
Yes	72	8.96	26	5.39
Driver drug use
No	728	90.55	456	94.61
Yes	76	9.45	26	5.39
Pedestrian characteristics
Pedestrian age
0–25 (young)	108	13.43	82	17.01
26–45 (young adult)	227	28.23	136	28.22
46–64 (middle aged)	261	32.46	131	27.18
65+ (senior)	208	25.87	133	27.59
Pedestrian gender
Female	516	64.18	322	66.80
Male	285	35.45	158	32.78
Unknown	3	0.37	2	0.41
Pedestrian location
Intersection-related	194	24.13	107	22.20
Non-intersection-related	608	75.62	375	77.80
Unknown	2	0.25	0	0.00
Pedestrian circumstances
Walking/standing/running	347	43.16	170	35.27
Vehicle-related	97	12.06	103	21.37
Motorist failed to yield	151	18.78	85	17.63
Pedestrian failed to yield	203	25.25	123	25.52
Unknown	6	0.75	1	0.21
Roadway characteristics
Functional class
Arterials	574	71.39	310	64.32
Collectors	122	15.17	73	15.15
Local	102	12.69	82	17.01
Unknown	6	0.75	17	3.53
Speed limit (mph)
0–25	93	11.57	81	16.80
30–45	422	52.49	223	46.27
50–65	223	27.74	140	29.05
70–85	43	5.35	17	3.53
Unknown	23	2.86	21	4.36
Number of lanes
One lane	7	0.87	9	1.87
Two lanes	430	53.48	280	58.09
More than two lanes	349	43.41	176	36.51
Unknown	18	2.24	17	3.53
Road division
Two-way undivided	499	62.06	319	66.18
Two-way divided	278	34.58	136	28.22
One-way	15	1.87	11	2.28
Unknown	12	1.49	16	3.32
Control
No control	579	72.01	368	76.35
Traffic signal	96	11.94	55	11.41
Traffic signs	50	6.22	24	4.98
Unknown	79	9.83	35	7.26
Crosswalk
None noted	653	81.22	390	80.91
Yes	147	18.28	92	19.09
Unknown	4	0.50	0	0.00
Sidewalk
None noted	528	65.67	317	65.77
Yes	270	33.58	162	33.61
Unknown	6	0.75	3	0.62
Environmental characteristics
Land use
Urban	612	76.12	357	74.07
Rural	186	23.13	108	22.41
Unknown	6	0.75	17	3.53
Weather
Clear	568	70.65	325	67.43
Rain	47	5.85	33	6.85
Snow/hail	11	1.37	12	2.49
Cloudy	132	16.42	90	18.67
Unknown	46	5.72	22	4.56
Lighting condition
Daylight	317	39.43	171	35.48
Dark–not lighted	231	28.73	154	31.95
Dark-lighted	221	27.49	134	27.80
Dusk/dawn	35	4.35	21	4.36
Unknown	0	0.00	2	0.41
Hour of the day
Late night (midnight–3:00)	99	12.31	64	13.28
Early morning (4:00–7:00)	113	14.05	70	14.52
Morning (8:00–11:00)	104	12.94	52	10.79
Midday (midday–15:00)	99	12.31	62	12.86
Evening (16:00–19:00)	181	22.51	110	22.82
Night (20:00–23:00)	208	25.87	122	25.31
Unknown	0	0.00	2	0.41
Day of the week
Weekend	206	25.62	138	28.63
Weekday	598	74.38	344	71.37
Vehicle characteristics
Vehicle classification
Passenger cars	384	47.76	198	41.08
SUVs/utility vehicles	186	23.13	99	20.54
Light trucks, vans, and buses	181	22.51	139	28.84
Large/heavy trucks	48	5.97	41	8.51
Motorcycles	0	0.00	1	0.21
Unknown	5	0.62	4	0.83
Vehicle model year
1980–1995	17	2.11	18	3.73
1996–2008	314	39.05	181	37.55
2009–2018	390	48.51	234	48.55
2019–2024	76	9.45	44	9.13
Unknown	7	0.87	5	1.04

Note: SUVs = sports utility vehicles.

Modeling Methodology

This study employed ARM, which is an unsupervised machine learning method that mainly identifies co-occurring items in a dataset. This method is commonly used in business ( 21 ), however, recently it has been used in diverse fields of research, including transportation engineering ( 22 , 23 ). The ARM method suits this study owing to its ability to uncover “if–then” patterns within multivariate datasets, without requiring a predefined target variable, unlike supervised machine learning techniques. Whereas other unsupervised methods, such as clustering, provide comparable benefits, they primarily group similar observations without explicitly revealing intervariable dependencies. ARM enables the identification of meaningful relationships and co-occurrence patterns among variables, making it more appropriate for examining the relationships among factors involved in distraction-related pedestrian fatalities with the use of an a priori algorithm.

A brief review of the methodology is provided below.

Let A = {a₁, a₂a₃, …} be a set of N crash attributes called variables (set of factors contributing for each pedestrian crash record) and D = {d₁, d₂, d₃, …} be a database of pedestrian crash information such that each crash record in D has a unique identification number, t_a, such that each t_a corresponds to a subset on A satisfying t_a⊆A ( 9 ).

The association rule is written as antecedent (P) → consequent (Q) where from the study, P is crash variables in relation to the crash and Q the distraction type leading to pedestrian fatalities. These rules can be filtered by three critical parameters: support (S), confidence (C), and lift (L).

Support refers to how often the antecedent (P) and consequent (Q) of a given rule appear in the entire dataset. This is calculated as shown in Equation 1.

Support (P \to Q) = \frac{P \cap Q}{T}

(1)

where

T = crash frequency,

P = number of observations with P,

Q = number of occurrences with Q, and

$P \cap$ Q = number of crashes with P and Q together.

Confidence measures the reliability of the inference of a generated rule. Higher confidence for P → Q indicates that the presence of Q is highly visible when having P ( 24 ). Equation 2 presents the calculation of confidence.

Confidence (P \to Q) = \frac{Support (P \to Q)}{Support (P)}

(2)

Lift signifies the correlation between antecedent and consequent by comparing their observed co-occurrence with the expected co-occurrence if they are statistically independent. A lift value greater than 1 means a positive correlation between the antecedent and the consequent, and vice versa for a value less than 1 ( 25 – 27 ). The lift can be computed as given in Equation 3.

Lift (P \to Q) = \frac{Support (P \to Q)}{Support (P) \times Support (Q)}

(3)

A rule with a single antecedent and a single consequent is defined as a two-product rule; similarly, a rule with two antecedents and a single consequent or one antecedent and two consequents is defined as a three-product rule.

Results and Discussion

To develop significant rules using ARM, it is crucial to define a minimum threshold of support and confidence, otherwise, the algorithm might produce millions of rules. The minimum support and confidence thresholds are typically established through iterative testing ( 24 ). In this study, these values were adjusted repeatedly to generate a set of rules that would be sufficiently informative for explaining the relationships between antecedents and consequents. Although varying the thresholds changes the number of rules produced, the overall patterns and key findings generally remain consistent. The value of lift provides a strength of the relationship between antecedents and consequents. In this study, the minimum threshold for the lift value was taken as 1.1, consistent with the recommended value ( 25 – 27 ).

For ease of interpretation, the study was limited to three-product rules. To identify and explain any pattern associated with generated rules, a rule “ID” was designed. From the selected variables, ARM was applied to two different cases (Case 1: distraction type = in-vehicle distraction, Case 2: distraction type = external distraction). The rules containing “unknown values” were excluded in the tables owing to the inability to extract useful information from them. The study used Python software for analysis and Stata software for cleaning.

Case 1: Distraction Type—In-Vehicle Distraction

The consequent was set to “in-vehicle distraction” as the right-hand side (RHS) to develop the association rules in this case. The minimum threshold of support and confidence was determined to be 0.5% and 30%, respectively. The Top 20 of the 412 rules determined are presented in descending order of their lift value in Table 3. The Top 10 rules were presented in a balloon plot in which the size of the balloon indicates the support values and the color intensity indicates the lift values (see Figure 3).

Table 3.

Top 20 Association Rules for In-Vehicle Distractions

ID	Antecedents (left-hand side)	Support, %	Confidence, %	Lift
R1	driver_age = 14–25 (young)	18.35	70.87	1.60
R2	speed = 70–85 (high)	3.34	71.67	1.60
R3	driver_drugs = yes	5.91	74.51	1.60
R4	driver_alcohol = yes	5.60	73.47	1.60
R5	roadclass = arterials, speed = 70–85 (high)	3.34	72.88	1.49
R6	roadclass = arterials, light = daylight	15.47	68.86	1.48
R7	roadclass = arterials, hour_group = morning (8:00–11:00)	5.44	72.16	1.48
R8	roadclass = arterials, driver_age = 14–25 (young)	12.36	72.60	1.48
R9	roadclass = arterials, driver_age = 26–45 (young adult)	22.24	71.14	1.47
R10	roadclass = arterials, driver_gender = female	14.07	69.62	1.44
R11	Roadclass = arterials, ped_circumstances_walking/standing/running	17.81	69.18	1.40
R12	roadclass = arterials, control_group = traffic signs	2.26	70.73	1.40
R13	roadclass = arterials, mod_year_class = 2019–2024	4.67	68.97	1.40
R14	roadclass = arterials, vehicle_group = SUVs	10.96	70.15	1.39
R15	roadclass = collectors, speed = 30–45 (moderate)	5.29	72.34	1.37
R16	roadclass = collectors, light = daylight	4.35	69.14	1.36
R17	roadclass = collectors, hour_group = evening (16:00–19:00)	2.49	74.42	1.36
R18	roadclass = collectors, driver_age = 14–25 (young)	3.19	71.93	1.35
R19	roadclass = collectors, ped_age = 46–64 (middle aged)	2.95	69.09	1.35
R20	roadclass = collectors, ped_circumstances = motorist failed to yield	1.87	70.59	1.33

Note: SUVs = sports utility vehicles; R = Rule.

Interpretation: R1 indicates that drivers aged 14 to 25 years were strongly linked to fatal pedestrian crashes involving in-vehicle distractions (S = 18.35%, C = 70.87%, L = 1.6). This age group accounted for 18.35% of such crashes, and among crashes involving these drivers, 70.87% were caused by in-vehicle distractions. Overall, the likelihood of fatal pedestrian crashes related to in-vehicle distractions among this age group was 1.6 times higher than in the overall dataset.

Driver Characteristics

Driver Age

Drivers aged 14 to 25 years also appeared in Rule R8, whereas those aged 26 to 45 years were represented in Rule R9. The frequent presence of young drivers in the rules suggests that they are more prone to in-vehicle distractions, such as using electronic devices or talking to passengers, which are common findings in previous studies ( 10 , 14 , 22 ). However, this finding contrasts with another study stating that drivers aged 18 to 44 years were less likely to be involved in crashes caused by in-vehicle distractions ( 7 ).

Driver Gender

In Rule R10, female drivers were associated with in-vehicle distractions. Although women are generally considered to be more cautious behind the wheel, the findings indicated a different pattern. These results are consistent with a study that found female drivers were more likely to be distracted by cellphone use ( 15 ).

Driving Under the Influence

Alcohol and drug use were prominent in Rules R3 and R4, likely reflecting the prevalence of young drivers who are more susceptible to such behaviors. Similar patterns were reported by studies on in-vehicle distraction-related crashes ( 7 , 8 ).

Vehicle Characteristics

Vehicle Model Year

The results for vehicle model year (R13) showed that newer vehicle models (2019–2024) were more frequently linked to in-vehicle distractions; however, a study by Kutela et al. found that newer vehicles were less likely to be involved in such incidents ( 7 ).

The relationship between new vehicle models and in-vehicle distractions produced mixed effects. Technologies like voice controls and driver-monitoring systems can reduce distraction, whereas features like complex infotainment systems, large touchscreens, and advanced driver-assistance systems can create a false sense of security, reducing driver awareness. A previous study noted that convenience features such as Bluetooth and Wi-Fi increase distractions; however, safety features such as lane departure warning and automatic braking help to reduce vehicle distractions ( 28 ). Therefore, more research is needed on how to incorporate these technologies and reduce distractions.

Vehicle Classification

The results also indicated that SUVs were frequently involved in crashes associated with in-vehicle distractions (R14). These vehicles typically feature larger cabin spaces, multiple storage areas, and easy access to personal items. This interior layout, combined with a generally higher perception of safety, may encourage more relaxed levels of attention and a greater willingness to multitask, thereby increasing the likelihood of engaging in in-vehicle distractions. In contrast, Islam concluded that inattentive driving is a prominent issue for SUV drivers but was not attributed to a specific distraction source ( 13 ).

Road Characteristics

Road Classification

Most crashes occurred along arterial roads (R5 to R14) and those with speed limit of 70 to 85 mph (R2, R5). Higher speeds and longer trips increased the tendency for drivers to use electronic devices, contributing significantly to in-vehicle distractions. Recent studies revealed similar findings: in-vehicle distractions are more common on high-speed and arterial roads ( 7 , 15 ). Another study reported that individuals driving at high speeds were more likely to be distracted by using cellphones ( 29 ). Collector roads were also dominant in the rules (R15 to R20) with the moderate speed limit (R15). The moderate-speed environment for collector roads may also create a false sense of safety, thus prompting drivers to use their electronic devices.

Although high-speed corridors such as expressways prohibit pedestrian activity, studies show that pedestrians continue to be present as a result of vehicle breakdowns, crossing attempts, or alcohol impairment ( 30 ). In these situations, the absence of pedestrian facilities combined with high operating speeds creates an environment where even a momentary lapse in drivers’ attention could result in a fatal crash.

Traffic Control Device

Another key factor identified was the presence of traffic signals (R12), indicating that signalized intersections are high-risk areas for pedestrian safety. This is seen when drivers often become distracted by mobile phones while they are stopped at red lights, reducing their attention once the signal changes ( 31 ). A previous study reported that traffic signs, as opposed to signals, are associated with a lower likelihood of severe crashes involving in-vehicle distractions ( 32 ).

Environmental Characteristics

Most crashes occurred during daylight (R6), particularly in the morning around 8:00 to 11:00 (R7) and in the evening around 16:00 to 19:00 (R17). This pattern aligns with pedestrian activity trends, as most people prefer walking during the day when visibility is higher and travel demand is greater.

Pedestrian Characteristics

The results indicated that, in many pedestrian-involved crashes, pedestrians were engaged in activities such as walking, standing, or running (R11). Some crashes also resulted from motorists failing to yield the right of way to pedestrians (R20), which is a common contributing factor in pedestrian collisions. The pedestrian age prevailing in the rules (R19) was 46 to 64 years, suggesting that middle-aged and older adults were more frequently involved in these incidents.

Case 2: Distraction Type—External Distraction

The consequent was set to “external distraction” as the RHS to develop the association rules in this case. The minimum threshold of support and confidence was determined at 1% and 40%, respectively. The Top 20 of the 459 rules determined are presented in descending order of their lift value in Table 4. The Top 10 rules are presented in a balloon plot in which the size of the balloon indicates the support values, whereas the color intensity indicates the lift values (Figure 4).

Table 4.

Top 20 Association Rules for External Distractions

ID	Antecedents (left-hand side)	Support, %	Confidence, %	Lift
R1	roadclass = local	6.38	44.57	2.21
R2	speed = 0–25 (low)	6.30	46.55	2.16
R3	driver_age = 46–64 (middle aged)	11.28	52.35	2.13
R4	driver_age = 65+ (senior)	3.81	62.82	2.11
R5	ped_age = 0–25	6.38	43.16	2.06
R6	weather_group = rain	2.57	41.25	2.04
R7	ped_circumstances = vehicle-control-related	8.01	51.50	2.03
R8	mod_year_class = 1980–1995	1.40	51.43	1.98
R9	vehicle_group = large/heavy trucks	3.19	46.07	1.97
R10	vehicle_group = light trucks, vans, and buses	10.81	43.44	1.97
R11	roadclass = arterials, driver_age = 46–64 (middle aged)	7.93	50.00	1.97
R12	roadclass = arterials, driver_age = 65+ (senior)	2.33	60.00	1.97
R13	roadclass = arterials, weather_group = rain	1.87	42.11	1.94
R14	roadclass =arterials, ped_circumstances = vehicle-control-related	4.74	43.57	1.93
R15	roadclass = collectors, speed = 50–65 (moderately high)	2.57	48.53	1.93
R16	roadclass = collectors, light = dark unlighted	2.33	42.86	1.93
R17	roadclass = collectors, crosswalk_present = yes	1.17	44.12	1.89
R18	roadclass = collectors, hour_group = early morning (4:00–7:00)	1.09	43.75	1.89
R19	roadclass = collectors, ped_gender = male	3.81	41.53	1.89
R20	roadclass = collectors, weather_group = cloudy	1.48	46.34	1.89

Interpretation: R1 indicates that fatal pedestrian crashes involving external distractions were strongly associated with local roads (S = 6.38%, C = 44.57%, L = 2.21). Specifically, 6.38% of such crashes occurred on local roads, and 44.57% of all fatal pedestrian crashes on local roads were the result of external distractions. Overall, the likelihood of these crashes on local roads was 2.21 times higher than in the overall dataset.

Driver Characteristics

Driver Age

Drivers aged 46 to 64 were prominent in Rules R3 and R11, with similar patterns observed among drivers aged 65 years and older (R4, R12). In contrast to in-vehicle distractions, these middle-aged and older drivers were more frequently associated with external distractions. Older drivers may be more inclined to visually scan or observe their surroundings, increasing their susceptibility to external stimuli. This trend is consistent with another study that found older drivers exhibited a greater likelihood of external distraction ( 15 ).

Conversely, findings based on naturalistic driving data indicate that teen drivers aged 16 to 20 experienced the highest crash risk from external distractions ( 14 ). This discrepancy may be attributed to methodological differences, specifically the use of naturalistic driving data versus crash data, as well as potential varying definitions and classifications of distraction across the studies. These inconsistencies warrant further research to clarify age-related patterns in external distraction.

Vehicle Characteristics

Vehicle Model Year

The study indicated that crashes predominantly involved older vehicles, specifically those manufactured between 1980 and 1995 (R8). Vehicles from this period typically incorporated only basic safety features, such as airbags, antilock braking systems, and lacked the infotainment technologies found in modern cars. With fewer in-vehicle features competing for driver attention, occupants may be more inclined to focus on the external driving environment, which could increase susceptibility to external distractions. However, a previous study suggested that drivers of older vehicles were significantly involved in crashes caused by reaching for fallen objects and cellphone use in the vehicle ( 7 ).

Vehicle Classification

Within this group, the crashes mostly involved large or heavy trucks (R9) or light trucks and buses (R10). Drivers in larger vehicles may benefit from a higher driver position, allowing for better observation of external factors compared with those in smaller vehicles. This observation contrasts with the findings of another study, which reported that external distractions were more significant in passenger cars owing to their lower seating position and smaller field of view when observing external objects, which increases susceptibility to distraction-related crashes ( 13 ).

Road Characteristics

Road Classification

Most of the crashes occurred along local roads, as observed in rule R1. The speed limit of 0 to 25 mph (R2) was observed, which is a common speed limit in residential areas. At low speeds, the drivers were more tempted to observe the surroundings and change focus as a result of easier driving conditions compared with roads with higher speeds. These findings were similar to those of Huisingh et al., who stated that external distractions are more common in local streets ( 15 ).

However, in some cases the crashes were also dominant in arterial roads (R11 to R14). Arterial roads carry more traffic and operate at higher speeds than local streets, creating a more demanding driving environment. Drivers are required to engage in frequent lane changes, merging, and diverging movements, which increases their mental workload. The presence of external features such as advertisements compound the visual distraction. Brome et al. demonstrated that roadside visual elements such as animated and transitioning billboards substantially impair driver attention ( 11 ).

Similarly, crashes were also prominent along collector roads (R15 to R20) with the moderate (50 to 65 mph) speed limit in R15. These roads function as intermediate links between local streets and arterials. This environment may encourage momentary lapses in attention or engagement in secondary tasks, which increases the likelihood of glancing at the roadside or interacting with in-vehicle devices.

Presence of Crosswalks

The crashes were dominant in areas with crosswalks (R17). Although crosswalks are meant to improve safety, pedestrians may assume that drivers will always yield. At the same time, distracted drivers may fail to notice pedestrians entering the crossing. This mismatch in expectations increases the likelihood of crashes.

Environmental Characteristics

Most crashes related to external distractions occurred in dark unlighted (R16) and early morning periods (R18). The weather conditions highlighted were rain (R6, R13) and cloudy (R20). These rules were associated with limited visibility along the road from insufficient or no light or weather conditions. Similar visibility-related risk patterns have been widely reported in research related to pedestrian crashes. Darkness with or without lighting conditions are key factors in fatal pedestrian crashes ( 22 , 33 ).

Pedestrian Characteristics

The pedestrian age identified in the rules was 0 to 25 years (R5), the pedestrian gender was male (R19). The analysis additionally identified vehicle-control-related issues (R6) as the dominant factor contributing to these crashes. Younger pedestrians often engage in riskier crossing behaviors. The analysis also showed that vehicle control issues (R6), such as failing to maintain lane position, were main contributing factors. This combination of unsafe pedestrian behavior and reduced driver control creates the likelihood of crash occurrence.

Comparative Analysis and Practical Implications

The comparative analysis between in-vehicle and external distraction-related pedestrian crashes revealed distinct patterns with significant practical implications, as explained below.

Driver Characteristics

In-vehicle distractions were associated with younger drivers aged 14 to 25 and 26 to 45, often combined with alcohol or drug use. External distractions were concentrated among middle-aged and older drivers (46 to 65+ years).

These findings highlighted that distracted driving was strongly age-dependent rather than a uniform risk across all drivers. Younger drivers require targeted interventions to reduce in-vehicle distractions. In contrast, middle-aged and older drivers would benefit from strategies aimed at minimizing external distractions in the driving environment.

Vehicle Characteristics

Newer vehicles models (2019 to 2024) and SUVs were frequently involved in in-vehicle distraction crashes, whereas older vehicles models (1980 to 1995) and larger vehicle types such as trucks, vans, and buses were more prominent in external distraction crashes.

These results emphasized the evolving role of vehicle design in shaping distraction-related crash patterns. Newer vehicles and SUVs may inadvertently increase opportunities for in-vehicle distraction. This addresses the need for manufacturers to prioritize designs that minimize distractions while maintaining comfort and safety.

Conversely, older vehicles appeared to be more prone to external distraction crashes, possibly because their simpler interiors direct driver attention outward. Drivers of all vehicle types should be educated on the specific distraction risks associated with their vehicles and encouraged to adopt safer driving behaviors.

Roadway Characteristics

In-vehicle crashes occurred mostly on arterial roads with 70- to 85-mph speed limits and on collector roads with 30- to 45-mph speed limits. External distraction crashes were more common on local roads with 0- to 25-mph speed limits and on collector roads with 50- to 65-mph speed limits.

Distraction risks differed across roadway types and speeds. High-speed arterials and moderate-speed collector roads were more prone to in-vehicle distractions, whereas local and lower-speed collector roads observed more external distractions. Since changing speed limits is challenging owing to road functional classification, targeted driver education and awareness campaigns are essential to help drivers adjust their attention according to the roadway environment.

Environmental Characteristics

Daylight and morning/evening peaks were characteristic of in-vehicle distraction crashes. Dark, unlighted, rainy, or cloudy conditions dominated external distraction crashes.

This highlights the need to improve roadway lighting as visibility could help reduce external distraction crashes in dark or poor weather. Drivers should also be reminded to stay attentive during daytime peak hours when in-vehicle distractions are more common.

Strategies to Mitigate Distracted-Driving Crashes

The findings suggest that mitigating distracted driving will require a multifaceted approach that extends well beyond traditional enforcement strategies. Because existing laws are limited in their ability to address externally induced visual and cognitive distractions, practical interventions should emphasize roadway design improvements, such as simplifying roadside environments, regulating the placement and luminance of digital billboards, and enhancing pedestrian and work-zone visibility.

Technological advancements, including in-vehicle driver-monitoring systems and advanced driver-assistance features, could further help detect and counteract lapses in attention. Equally important is sustained public education aimed at raising awareness of both device-related and environmental distractions.

Conclusion and Recommendations

Using FARS data, the study used ARM to analyze patterns between different types of distractions and fatal pedestrian crashes in the United States over 8 years. The study was significant in producing patterns related to both in-vehicle and external distractions.

The following are the conclusions of the study:

Driver distraction patterns differ starkly by source, with in-vehicle distractions predominantly involving younger drivers, newer vehicles, and higher-speed arterial roads, whereas external distractions are more common among middle-aged and older drivers on low-speed local and collector roads.

In-vehicle distraction crashes are closely associated with risk-enhancing behaviors, including alcohol/drug use, use of electronic devices, and driving SUVs or late-model vehicles equipped with complex infotainment systems.

External distraction crashes are strongly tied to environmental and roadway visibility factors, occurring more often in dark, unlit, rainy, or cloudy conditions, and involving older vehicles, large trucks, vans, and buses.

Effective mitigation requires a comprehensive strategy that combines stronger enforcement of device-related distraction laws with roadway design improvements, visibility enhancements, and adoption of driver-monitoring and safety technologies to address both in-vehicle and external distractions.

The following are the recommendations derived from this study:

Strengthening primary enforcement of distracted-driving laws. According to the Governors Highway Safety Association, all but six states have primary enforcement of these laws ( 34 ). Adopting stricter primary enforcement policies would enable officers to issue penalties directly for texting and other distraction-related violations.

Development of infrastructural solutions. These include using CCTV systems and intelligent transportation systems, particularly vehicle-to-everything communication, which allows vehicles to interact with other vehicles, infrastructure, pedestrians, and networks to enhance awareness and safety.

Integration of real-time driver-monitoring systems for safer vehicles. Technologies such as eye-tracking sensors, facial recognition, and steering behavior analysis can detect signs of inattention and could trigger immediate alerts or corrective actions. This approach would complement enforcement-, infrastructure-based-, and education solutions. To address distracted driving that involves drivers under the influence, ignition interlocks in vehicles could also be used.

Improvement of road signage, fencing to promote safe roads, and providing roadside assistance for vehicles experiencing mechanical problems on high-speed corridors. The adoption of traffic calming measures in residential areas could help to reduce the risk of distraction-related crashes.

Visibility enhancements in dark or low-visibility areas to help drivers better identify pedestrians and avoid externally triggered distractions.

Age-specific education campaigns to address the distinct distraction risks among drivers.

Limitations and Further Research Suggestions

The study has several limitations that should be acknowledged. This study is based on crash data from the United States, its findings may therefore not be generalizable to countries with different road safety cultures, infrastructural conditions, or distracted-driving policies. In addition, more research is needed to better understand the mixed effects of vehicle type, vehicle model year, and roadway characteristics across different distraction sources. Moreover, complementing crash data with observational and naturalistic driving studies would provide deeper insight into driver behavior and the contextual conditions that elicit specific distraction types.

Footnotes

Acknowledgements

The authors gratefully acknowledge the use of data from the Fatality Analysis Reporting System, maintained by the National Highway Traffic Safety Administration.

Author Contributions

The authors confirm contribution to the paper as follows: study conception and design: U. Bitaliho, V. Kwigizile; data collection: U. Bitaliho, S. Mwende, N. Novat; analysis and interpretation of results: U. Bitaliho, S. Mwende, N. Novat; draft manuscript preparation: U. Bitaliho, V. Kwigizile, J. Oh. All authors reviewed the results and approved the final version of the manuscript.

Declaration of Conflicting Interests

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

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

ORCID iDs

Upendo Gerald Bitaliho

Norran Kakama Novat

Valerian Kwigizile

Jun- Seok Oh

Data Accessibility Statement

The datasets generated and/or analyzed during this study are publicly available and accessible through the Fatality Analysis Reporting System repository at: .

The opinions, findings, and conclusions expressed in this paper are those of the authors.

References

Qin

(Richard) Li

Chen

Andi Bill

M. S.

Noyce

D. A.

Understanding Driver Distractions in Fatal Crashes: An Exploratory Empirical Analysis. Journal of Safety Research, Vol. 69, 2019, pp. 23–31. https://doi.org/10.1016/j.jsr.2019.01.004

National Center for Statistics and Analysis. Distracted Driving in 2023 (Research Note. Report No. DOT HS 813 703). 2025. https://crashstats.nhtsa.dot.gov/Api/Public/Publication/813703

Stimpson

Wilson

Muelleman

Fatalities of Pedestrians, Bicycle Riders and Motorists Due to Distracted Driving Motor Vehicle Crashes in the U.S, 2005–2010. Public Health Reports, Vol. 128, No. 6, 2013, pp. 436–442. https://doi.org/10.1177/003335491312800603

Sajid Hasan

Patel

Alfaris

Jalayer

Identifying Distracted-Driving Events from on-Road Observations Using a Moving Vehicle: A Case Study in New Jersey. Accident Analysis and Prevention, Vol. 177, 2022. https://doi.org/10.1016/j.aap.2022.106827

Khan

N. A.

Habib

M. A.

Exploring the Impacts of Built Environment on Pedestrian Injury Severity Involving Distracted Driving. Journal of Safety Research, Vol. 80, 2022, pp. 97–108. https://doi.org/10.1016/j.jsr.2021.11.001

Divekar

Pradhan

A. K.

Pollatsek

Fisher

D. L.

Effect of External Distractions Behavior and Vehicle Control of Novice and Experienced Drivers Evaluated. Transportation Research Record: Journal of the Transportation Research Board, 2012. 2321(1): 15–22. https://doi.org/10.3141/2321-03

Kutela

Kidando

Kitali

A. E.

Mwende

Examining In-Vehicle Distraction Sources in Relation to Crashes Using a Bayesian Multinomial Logit Model. 2023. DOI: 10.53136/97912218091901

Hasan

A. S.

Orvin

M. M.

Jalayer

Heitmann

Weiss

Analysis of Distracted Driving Crashes in New Jersey Using Mixed Logit Model. Journal of Safety Research, Vol. 81, 2022, pp. 166–174. https://doi.org/10.1016/j.jsr.2022.02.008

Rahman

M. A.

Sun

Shan

Investigating Characteristics of Cellphone Distraction with Significance Tests and Association Rule Mining. IATSS Research, Vol. 45, No. 2, 2021, pp. 198–209. https://doi.org/10.1016/j.iatssr.2020.09.001

10.

Choudhary

Velaga

N. R.

Effects of Phone Use on Driving Performance: A Comparative Analysis of Young and Professional Drivers. Safety Science, Vol. 111, 2019, pp. 179–187. https://doi.org/10.1016/j.ssci.2018.07.009

11.

Brome

Awad

Moacdieh

N. M.

Roadside Digital Billboard Advertisements: Effects of Static, Transitioning, and Animated Designs on Drivers’ Performance and Attention. Transportation Research Part F: Traffic Psychology and Behaviour, Vol. 83, 2021, pp. 226–237. https://doi.org/10.1016/j.trf.2021.10.013

12.

Sheykhfard

Haghighi

Driver Distraction by Digital Billboards? Structural Equation Modeling Based on Naturalistic Driving Study Data: A Case Study of Iran. Journal of Safety Research, Vol. 72, 2020, pp. 1–8. https://doi.org/10.1016/j.jsr.2019.11.002

13.

Islam

Unraveling the Differences in Distracted Driving Injury Severities in Passenger Car, Sport Utility Vehicle, Pickup Truck, and Minivan Crashes. Accident Analysis and Prevention, Vol. 196, 2024. https://doi.org/10.1016/j.aap.2023.107444

14.

Guo

Klauer

S. G.

Fang

Hankey

J. M.

Antin

J. F.

Perez

M. A.

Lee

S. E.

Dingus

T. A.

The Effects of Age on Crash Risk Associated with Driver Distraction. International Journal of Epidemiology, Vol. 46, No. 1, 2017, pp. 258–265. https://doi.org/10.1093/ije/dyw234

15.

Huisingh

Griffin

McGwin

The Prevalence of Distraction Among Passenger Vehicle Drivers: A Roadside Observational Approach. Traffic Injury Prevention, Vol. 16, No. 2, 2015, pp. 140–146. https://doi.org/10.1080/15389588.2014.916797

16.

Horberry

Anderson

Regan

M. A.

Triggs

T. J.

Brown

Driver Distraction: The Effects of Concurrent in-Vehicle Tasks, Road Environment Complexity and Age on Driving Performance. Accident Analysis and Prevention, Vol. 38, No. 1, 2006, pp. 185–191. https://doi.org/10.1016/j.aap.2005.09.007

17.

Das

Dutta

Avelar

Dixon

Sun

Jalayer

Supervised Association Rules Mining on Pedestrian Crashes in Urban Areas: Identifying Patterns for Appropriate Countermeasures. International Journal of Urban Sciences, Vol. 23, No. 1, 2019, pp. 30–48. https://doi.org/10.1080/12265934.2018.1431146

18.

Rudisill

T. M.

Barbee

L. O.

Hendricks

Characteristics of Fatal, Pedestrian-Involved, Motor Vehicle Crashes in West Virginia: A Cross-Sectional and Spatial Analysis. International Journal of Environmental Research and Public Health, Vol. 20, No. 7, 2023. https://doi.org/10.3390/ijerph20075251

19.

Novat

N. K.

Kwigizile

J. S.

Impact of Slip Lanes on Pedestrian Safety at Roundabouts Considering Autonomous Vehicles. Transportation Research Record: Journal of the Transportation Research Board, 2025. 2679(3): 112–126. https://doi.org/10.1177/03611981241275562

20.

Deka

Brown

What Do Planning Professionals, Police, and Pedestrians in General Think about Distracted Driving and Walking?

Transportation Research Record: Journal of the Transportation Research Board, 2016. 2582: 42–50. https://doi.org/10.3141/2582-06

21.

Agrawal

Imieliński

Swami

Mining Association Rules between Sets of Items in Large Databases. 1993. https://doi.org/10.1145/170036.170072

22.

Hossain

Sun

Thapa

Codjoe

Applying Association Rules Mining to Investigate Pedestrian Fatal and Injury Crash Patterns Under Different Lighting Conditions. Transportation Research Record: Journal of the Transportation Research Board, 2022. 2676(6): 659–672. https://doi.org/10.1177/03611981221076120

23.

Xing

Chen

Zhang

Xie

Zhu

Understanding Distracted Driving Patterns of Ride-Hailing Drivers from Multi-Source Data: Applying Association Rule Mining. Journal of Transportation Safety and Security, Vol. 16, No. 4, 2024, pp. 390–420. https://doi.org/10.1080/19439962.2023.2221204

24.

Das

Kong

Tsapakis

Hit and Run Crash Analysis Using Association Rules Mining. Journal of Transportation Safety and Security, Vol. 13, No. 2, 2021, pp. 123–142. https://doi.org/10.1080/19439962.2019.1611682

25.

Hahsler

Grün

Hornik

Wien

Arules-A Computational Environment for Mining Association Rules and Frequent Item Sets. 2005. DOI: 10.18637/jss.v014.i15

26.

Han

Kamber

Pei

Data Mining. Concepts and Techniques, 3rd ed. (The Morgan Kaufmann Series in Data Management Systems). Morgan Kaufmann Publishers, Waltham, MA, 2011.

27.

Tan

P.-N.

Steinbach

Kumar

Introduction to Data Mining. Pearson, Pearson Addison-Wesley, Boston, MA, 2018.

28.

Schneider

E. M.

D’Ambrosio

L. A.

Lee

Coughlin

J. F.

Impacts of Advanced Vehicle Technologies and Risk Attitudes on Distracted Driving Behaviors. Transportation Research Record: Journal of the Transportation Research Board, 2024. 2678(11): 622–634. https://doi.org/10.1177/03611981241242079

29.

Zhao

Reimer

Mehler

D’Ambrosio

L. A.

Coughlin

J. F.

Self-Reported and Observed Risky Driving Behaviors among Frequent and Infrequent Cell Phone Users. Accident Analysis and Prevention, Vol. 61, 2013, pp. 71–77. https://doi.org/10.1016/j.aap.2012.07.019

30.

AAA Foundation for Traffic Safety. Pedestrian Fatalities on Interstate Highways, United States, 1993–2012. AAA Foundation for Traffic Safety, Washington, D.C., 2014. https://aaafoundation.org/research/pedestrian-fatalities-interstate-highways/

31.

Vera-López

J. D.

Pérez-Núñez

Híjar

Hidalgo-Solórzano

Lunnen

J. C.

Chandran

Hyder

A. A.

Distracted Driving: Mobile Phone Use While Driving in Three Mexican Cities. Injury Prevention, Vol. 19, No. 4, 2013, pp. 276–279. https://doi.org/10.1136/injuryprev-2012-040496

32.

Song

Meng

Temporal Analysis of Cellphone-Use-Involved Crash Injury Severities: Calling for Preventing Cellphone-Use-Involved Distracted Driving. Accident Analysis and Prevention, Vol. 169, 2022. https://doi.org/10.1016/j.aap.2022.106625

33.

Das

Tamakloe

Zubaidi

Obaid

Alnedawi

Fatal Pedestrian Crashes at Intersections: Trend Mining Using Association Rules. Accident Analysis and Prevention, Vol. 160, 2021. https://doi.org/10.1016/j.aap.2021.106306

34.

Governors Highway Safety Association. Distracted Driving: Issue Overview. Insurance Institute for Highway Safety (IIHS) and State Highway Safety Offices (SHSOs). https://www.ghsa.org/state-laws-issues/distracted-driving. Accessed November 29, 2025.