Pub Date : 2025-12-26DOI: 10.1016/j.trf.2025.103501
Emma Tivesten, Thomas Streubel, Mikael Ljung Aust
Advanced driver assistance systems that simultaneously perform lateral and longitudinal control reduce the need for driver input, potentially leading to driver disengagement. As these systems become more capable of performing most of the operational control, drivers tend to increase eyes off road, hands off wheel, and they may be unprepared to act in situations that exceed the system's capabilities.
In this study, we analyzed the behaviors of 54 participants using a level 2 system on a test track. Drivers were considered disengaged if they had a late or absent response to a conflict at the end of the drive, resulting in a crash or near-crash. Several behaviors were associated with increased risk of disengagement, including long off-path glances, frequent visual time-sharing, gaze concentration, lack of driver steering input, and hands off wheel during uneventful driving. In contrast, continuous engagement in steering appeared to promote driver engagement, even among participants who exhibited suboptimal gaze behavior.
These findings suggest that combining metrics of steering activity and gaze behavior provides a more comprehensive assessment of driver engagement. This insight can inform the design of driver monitoring and engagement strategies in level 2 driving systems.
{"title":"Eye, steering, and hands on wheel behaviors indicating driver engagement in level 2 driving","authors":"Emma Tivesten, Thomas Streubel, Mikael Ljung Aust","doi":"10.1016/j.trf.2025.103501","DOIUrl":"10.1016/j.trf.2025.103501","url":null,"abstract":"<div><div>Advanced driver assistance systems that simultaneously perform lateral and longitudinal control reduce the need for driver input, potentially leading to driver disengagement. As these systems become more capable of performing most of the operational control, drivers tend to increase eyes off road, hands off wheel, and they may be unprepared to act in situations that exceed the system's capabilities.</div><div>In this study, we analyzed the behaviors of 54 participants using a level 2 system on a test track. Drivers were considered disengaged if they had a late or absent response to a conflict at the end of the drive, resulting in a crash or near-crash. Several behaviors were associated with increased risk of disengagement, including long off-path glances, frequent visual time-sharing, gaze concentration, lack of driver steering input, and hands off wheel during uneventful driving. In contrast, continuous engagement in steering appeared to promote driver engagement, even among participants who exhibited suboptimal gaze behavior.</div><div>These findings suggest that combining metrics of steering activity and gaze behavior provides a more comprehensive assessment of driver engagement. This insight can inform the design of driver monitoring and engagement strategies in level 2 driving systems.</div></div>","PeriodicalId":48355,"journal":{"name":"Transportation Research Part F-Traffic Psychology and Behaviour","volume":"118 ","pages":"Article 103501"},"PeriodicalIF":4.4,"publicationDate":"2025-12-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145841823","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-25DOI: 10.1016/j.trf.2025.103487
Kun Wang , Rensu Zhou , Shuo Yang , Cheng Wang , Jing Liu , Lu Wei , LiangXu
Pedestrians' crossing the street while running a red light are a significant factor contributing to traffic accidents at intersections. Traditional models fail to capture the complex, multifactorial nonlinearities and interactions involved in this behavior due to their limited linear analytical power, while machine learning models suffer from interpretability issues. To address this, an analytical framework that combines data-driven machine learning algorithms with emerging interpretability techniques was proposed, aiming to reveal the complex, nonlinear effects and relative importance of factors influencing pedestrian red-light-crossing behavior. Empirical video data from five signalized intersections in Hefei, China were used to compare the modeling and prediction performance of four methods: logistic regression, K-nearest neighbors, support vector machine, and extreme gradient boosting (XGBoost). Shapley Additive Explanations (SHAP) and Accumulated Local Effects (ALE) were employed to evaluate the key factors influencing pedestrians' decisions to cross at a red light. The results show that the XGBoost model outperforms the other algorithms in capturing the complex relationships among influencing factors and accurately identifying red-light-running behavior. Quantitative analysis of feature importance reveals that traffic volume is the most influential predictor, followed by pedestrian walking speed, red-light duration, conformity behavior, and age. This study overcomes the linear constraints of traditional regression models and provides a theoretical foundation for optimizing traffic management and developing intelligent law enforcement strategies.
{"title":"Identification of factors influencing pedestrians' red-light crossing behavior based on interpretable machine learning","authors":"Kun Wang , Rensu Zhou , Shuo Yang , Cheng Wang , Jing Liu , Lu Wei , LiangXu","doi":"10.1016/j.trf.2025.103487","DOIUrl":"10.1016/j.trf.2025.103487","url":null,"abstract":"<div><div>Pedestrians' crossing the street while running a red light are a significant factor contributing to traffic accidents at intersections. Traditional models fail to capture the complex, multifactorial nonlinearities and interactions involved in this behavior due to their limited linear analytical power, while machine learning models suffer from interpretability issues. To address this, an analytical framework that combines data-driven machine learning algorithms with emerging interpretability techniques was proposed, aiming to reveal the complex, nonlinear effects and relative importance of factors influencing pedestrian red-light-crossing behavior. Empirical video data from five signalized intersections in Hefei, China were used to compare the modeling and prediction performance of four methods: logistic regression, K-nearest neighbors, support vector machine, and extreme gradient boosting (XGBoost). Shapley Additive Explanations (SHAP) and Accumulated Local Effects (ALE) were employed to evaluate the key factors influencing pedestrians' decisions to cross at a red light. The results show that the XGBoost model outperforms the other algorithms in capturing the complex relationships among influencing factors and accurately identifying red-light-running behavior. Quantitative analysis of feature importance reveals that traffic volume is the most influential predictor, followed by pedestrian walking speed, red-light duration, conformity behavior, and age. This study overcomes the linear constraints of traditional regression models and provides a theoretical foundation for optimizing traffic management and developing intelligent law enforcement strategies.</div></div>","PeriodicalId":48355,"journal":{"name":"Transportation Research Part F-Traffic Psychology and Behaviour","volume":"118 ","pages":"Article 103487"},"PeriodicalIF":4.4,"publicationDate":"2025-12-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145841824","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-24DOI: 10.1016/j.trf.2025.103483
Yanqing Yao , Wuyang Chen , Xiaoyu Zhuang , Chenyu Xu , Doyeon Lee , Xiaoou Yang , Jie Wang , Shijian He
When driving under time pressure, drivers often adjust their behavior to save travel time. Understanding these changes is important for evaluating their potential impacts on traffic safety and travel efficiency. This study examines how time pressure affects driving behavior, physiological responses, and travel time in complex urban road environments. Thirty-two young drivers completed two naturalistic driving tasks representing low and high time-pressure conditions on routes containing traffic-light-dense and non-dense segments. The driving behavior (speed, overspeeding frequency, and lane-changing frequency), physiological indicators (heart rate and skin conductance level), and travel time were recorded. The results show that time pressure led to significantly more assertive driving behaviors, with higher speeds and increased overspeeding and lane-changing frequency under high time pressure. By contrast, no statistically significant differences were observed in both the heart rate and skin conductance levels across the roadway segments. The effects of time pressure on the travel time were highly context dependent: no meaningful time savings occurred on traffic-light-dense segments, whereas small but measurable reductions were achieved on non-dense segments. These findings indicate that although time pressure reliably intensifies driving behavior, actual efficiency gains are limited and strongly constrained by roadway signal density. This evidence supports efforts in traffic safety policy and driver education to recalibrate drivers’ expectations regarding the effectiveness of assertive driving under time pressure.
{"title":"Driving under time pressure: Driver state and behavior changes with limited time savings in complex road networks - A naturalistic time-incentive study with young male drivers","authors":"Yanqing Yao , Wuyang Chen , Xiaoyu Zhuang , Chenyu Xu , Doyeon Lee , Xiaoou Yang , Jie Wang , Shijian He","doi":"10.1016/j.trf.2025.103483","DOIUrl":"10.1016/j.trf.2025.103483","url":null,"abstract":"<div><div>When driving under time pressure, drivers often adjust their behavior to save travel time. Understanding these changes is important for evaluating their potential impacts on traffic safety and travel efficiency. This study examines how time pressure affects driving behavior, physiological responses, and travel time in complex urban road environments. Thirty-two young drivers completed two naturalistic driving tasks representing low and high time-pressure conditions on routes containing traffic-light-dense and non-dense segments. The driving behavior (speed, overspeeding frequency, and lane-changing frequency), physiological indicators (heart rate and skin conductance level), and travel time were recorded. The results show that time pressure led to significantly more assertive driving behaviors, with higher speeds and increased overspeeding and lane-changing frequency under high time pressure. By contrast, no statistically significant differences were observed in both the heart rate and skin conductance levels across the roadway segments. The effects of time pressure on the travel time were highly context dependent: no meaningful time savings occurred on traffic-light-dense segments, whereas small but measurable reductions were achieved on non-dense segments. These findings indicate that although time pressure reliably intensifies driving behavior, actual efficiency gains are limited and strongly constrained by roadway signal density. This evidence supports efforts in traffic safety policy and driver education to recalibrate drivers’ expectations regarding the effectiveness of assertive driving under time pressure.</div></div>","PeriodicalId":48355,"journal":{"name":"Transportation Research Part F-Traffic Psychology and Behaviour","volume":"118 ","pages":"Article 103483"},"PeriodicalIF":4.4,"publicationDate":"2025-12-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145841793","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-20DOI: 10.1016/j.trf.2025.103484
Duy Quy Nguyen-Phuoc , Thao Nhi Ho-Mai , Thao Phuong Thi Nguyen , Nhat Dinh Quang Vo , Tuan Trong Luu , Diep Ngoc Su
Motorcycles dominate urban transport in Southeast Asia but significantly contribute to air pollution and health risks. Government efforts to promote electric motorcycles (EMs) have been extensive, yet consumer switch rates continue to lag behind expectations. Currently, AI plays a transformative role in the development, functionality, and optimisation of EMs; however, the influence of beliefs in AI benefits and trust in AI on EM switching intention has not been examined. This study addresses this gap by extending the Value-Belief-Norm theory with these two constructs, incorporating age and gender as moderators. Additionally, perceived value is treated as a second-order construct, consisting of five dimensions that collectively capture its multifaceted nature. Data from Vietnamese respondents were analysed using PLS-SEM. The results show that perceived value directly influences beliefs in EM benefits, AI benefits for EMs, and trust in AI technologies. Trust in AI and personal norms significantly shape the switching intention. This research provides actionable recommendations for policymakers to accelerate EM switching intention and contributes to advancing theoretical discussions on sustainable transport and AI integration in mobility solutions.
{"title":"Driving the shift to electric motorcycles: the role of AI trust and AI benefits","authors":"Duy Quy Nguyen-Phuoc , Thao Nhi Ho-Mai , Thao Phuong Thi Nguyen , Nhat Dinh Quang Vo , Tuan Trong Luu , Diep Ngoc Su","doi":"10.1016/j.trf.2025.103484","DOIUrl":"10.1016/j.trf.2025.103484","url":null,"abstract":"<div><div>Motorcycles dominate urban transport in Southeast Asia but significantly contribute to air pollution and health risks. Government efforts to promote electric motorcycles (EMs) have been extensive, yet consumer switch rates continue to lag behind expectations. Currently, AI plays a transformative role in the development, functionality, and optimisation of EMs; however, the influence of beliefs in AI benefits and trust in AI on EM switching intention has not been examined. This study addresses this gap by extending the Value-Belief-Norm theory with these two constructs, incorporating age and gender as moderators. Additionally, perceived value is treated as a second-order construct, consisting of five dimensions that collectively capture its multifaceted nature. Data from Vietnamese respondents were analysed using PLS-SEM. The results show that perceived value directly influences beliefs in EM benefits, AI benefits for EMs, and trust in AI technologies. Trust in AI and personal norms significantly shape the switching intention. This research provides actionable recommendations for policymakers to accelerate EM switching intention and contributes to advancing theoretical discussions on sustainable transport and AI integration in mobility solutions.</div></div>","PeriodicalId":48355,"journal":{"name":"Transportation Research Part F-Traffic Psychology and Behaviour","volume":"118 ","pages":"Article 103484"},"PeriodicalIF":4.4,"publicationDate":"2025-12-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145799803","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-20DOI: 10.1016/j.trf.2025.103493
Weiwei Wang , Zhiqiang Wen , Qizhao Peng , Zihao Zhang , Congge Shi , Ting Wei
Young novice drivers are prone to negative emotions in high-risk situations. These emotions consume limited cognitive resources and raise crash risk. Existing research has not systematically clarified the relationships among driving scenarios, emotion types, emotion intensity, and multi-channel cognitive workload. Accordingly, this study used a driving-simulator experiment to analyze these linkages. We recruited 144 Chinese young novice drivers and used pre-validated video clips to induce neutral, anger, fear, anxiety. Data were collected via the Self-Assessment Manikin (SAM), the Visual-Auditory-Cognitive-Psychomotor (VACP) workload model, and semi-structured interviews. The results showed that: (1) Negative emotions significantly increased cognitive workload in young novice drivers. Anger and fear causd significant instantaneous workload fluctuations, whereas anxiety yielded the highest mean workload. (2) Distinct negative emotions were triggered by specific driving scenarios, which have different stressors (such as security threat, time pressure and environmental complexity). The potential outcome brought by these situational stressors affect the intensity of emotion. (3) Emotion intensity was positively associated with workload level. High-arousal emotions more likely to increase demands on visual, cognitive, and psychomotor resources. Within a unified paradigm, this study delineates the pathway linking driving scenarios, emotion types, emotion intensity, and multi-channel workload. The findings provide evidence for in-vehicle emotion monitoring and environmental-adaptive interventions.
{"title":"Investigating the negative emotional intensity and cognitive workload levels of young Novice drivers in different high-risk driving scenarios: A simulated driving study","authors":"Weiwei Wang , Zhiqiang Wen , Qizhao Peng , Zihao Zhang , Congge Shi , Ting Wei","doi":"10.1016/j.trf.2025.103493","DOIUrl":"10.1016/j.trf.2025.103493","url":null,"abstract":"<div><div>Young novice drivers are prone to negative emotions in high-risk situations. These emotions consume limited cognitive resources and raise crash risk. Existing research has not systematically clarified the relationships among driving scenarios, emotion types, emotion intensity, and multi-channel cognitive workload. Accordingly, this study used a driving-simulator experiment to analyze these linkages. We recruited 144 Chinese young novice drivers and used pre-validated video clips to induce neutral, anger, fear, anxiety. Data were collected <em>via</em> the Self-Assessment Manikin (SAM), the Visual-Auditory-Cognitive-Psychomotor (VACP) workload model, and semi-structured interviews. The results showed that: (1) Negative emotions significantly increased cognitive workload in young novice drivers. Anger and fear causd significant instantaneous workload fluctuations, whereas anxiety yielded the highest mean workload. (2) Distinct negative emotions were triggered by specific driving scenarios, which have different stressors (such as security threat, time pressure and environmental complexity). The potential outcome brought by these situational stressors affect the intensity of emotion. (3) Emotion intensity was positively associated with workload level. High-arousal emotions more likely to increase demands on visual, cognitive, and psychomotor resources. Within a unified paradigm, this study delineates the pathway linking driving scenarios, emotion types, emotion intensity, and multi-channel workload. The findings provide evidence for in-vehicle emotion monitoring and environmental-adaptive interventions.</div></div>","PeriodicalId":48355,"journal":{"name":"Transportation Research Part F-Traffic Psychology and Behaviour","volume":"118 ","pages":"Article 103493"},"PeriodicalIF":4.4,"publicationDate":"2025-12-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145799802","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-17DOI: 10.1016/j.trf.2025.103480
Ina Koniakowsky , Yannick Forster , Frederik Naujoks , Josef F. Krems , Andreas Keinath
Driver monitoring systems (DMS) represent a camera-based countermeasure for visual distraction that detect distracted drivers in real-time and subsequently prompt them to look back on the road. However, the effectiveness of DMS in reducing distraction is still being debated, with studies yielding inconsistent results. A correct understanding of a technological system is a key determinant for its effectiveness in terms of enhancing safety. Therefore, addressing a previously unexplored factor in DMS research, this study investigated how drivers' explicit knowledge about DMS influences system effectiveness in reducing visual distraction. Previous studies showed that drivers' understanding of DMS is incomplete if drivers are not instructed. Therefore, in this study, the drivers' explicit knowledge of DMS was systematically manipulated between participants by providing verbal instructions prior to driving. Three experimental conditions were compared: drivers with explicit knowledge of DMS, drivers without knowledge, and a control group with an inactive DMS (between factor). Glance behavior was compared between the first and repeated interaction with the secondary task to assess visual distraction (within factor). Explicit knowledge significantly reduced the number of glances exceeding 2 s, indicating reduced visual distraction. This effect was prevalent, even after repeated interaction with the DMS. Importantly, mere activation of DMS without instruction did not affect glance behavior. Findings highlight the significant role of explicit knowledge in system effectiveness. The present work contributes to the field of DMS research by investigating drivers' mental model of DMS and deriving methodological and practical implications.
{"title":"Explicit knowledge of driver monitoring systems changes their effectiveness in reducing visual distraction","authors":"Ina Koniakowsky , Yannick Forster , Frederik Naujoks , Josef F. Krems , Andreas Keinath","doi":"10.1016/j.trf.2025.103480","DOIUrl":"10.1016/j.trf.2025.103480","url":null,"abstract":"<div><div>Driver monitoring systems (DMS) represent a camera-based countermeasure for visual distraction that detect distracted drivers in real-time and subsequently prompt them to look back on the road. However, the effectiveness of DMS in reducing distraction is still being debated, with studies yielding inconsistent results. A correct understanding of a technological system is a key determinant for its effectiveness in terms of enhancing safety. Therefore, addressing a previously unexplored factor in DMS research, this study investigated how drivers' explicit knowledge about DMS influences system effectiveness in reducing visual distraction. Previous studies showed that drivers' understanding of DMS is incomplete if drivers are not instructed. Therefore, in this study, the drivers' explicit knowledge of DMS was systematically manipulated between participants by providing verbal instructions prior to driving. Three experimental conditions were compared: drivers with explicit knowledge of DMS, drivers without knowledge, and a control group with an inactive DMS (between factor). Glance behavior was compared between the first and repeated interaction with the secondary task to assess visual distraction (within factor). Explicit knowledge significantly reduced the number of glances exceeding 2 s, indicating reduced visual distraction. This effect was prevalent, even after repeated interaction with the DMS. Importantly, mere activation of DMS without instruction did not affect glance behavior. Findings highlight the significant role of explicit knowledge in system effectiveness. The present work contributes to the field of DMS research by investigating drivers' mental model of DMS and deriving methodological and practical implications.</div></div>","PeriodicalId":48355,"journal":{"name":"Transportation Research Part F-Traffic Psychology and Behaviour","volume":"117 ","pages":"Article 103480"},"PeriodicalIF":4.4,"publicationDate":"2025-12-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145796916","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-17DOI: 10.1016/j.trf.2025.103482
Hyunchul Park, Taeho Oh, Jaehyuck Lim, Inhi Kim
While micromobility offers flexible solutions for first and last mile transportation, harmonizing with the existing public transportation, it raises substantial safety concerns, especially when overtaken by vehicles. Since the road infrastructure and regulations for micromobility have not kept pace with the rapid expansion, understanding the unique safety issues related to micromobility remains a critical task. This study aims to investigate the influence of cross-modal experience—drivers' direct experience with micromobility and road conditions on overtaking behavior towards micromobility, which includes e-scooters and bicycles, representatively. Using a VR-based simulator experiment, data were collected from 30 participants to analyze overtaking behavior. A linear mixed model was applied to assess how these factors impact driver's behaviors, such as lateral clearance, overtaking speed, and gaze fixation counts. The results indicate that drivers with cross-modal experience maintain wider lateral clearances when overtaking both e-scooters and bicycles, reflecting heightened awareness of micromobility safety needs. Additionally, poor road conditions make drivers increase lateral clearance and gaze fixation, emphasizing the importance of road infrastructure in ensuring safety. However, cross-modal experience did not significantly affect overtaking speed and road conditions for e-scooters. Conversely, a significant interaction with bicycles highlights the complexity of driving behavior and the need for comprehensive safety strategies. These findings support the development of educational programs promoting cross-modal experiences and targeted infrastructure improvements, such as dedicated e-scooter lanes, to enhance road safety for all users.
{"title":"Micro mobility safety challenges: a study on drivers overtaking bicycles and E-scooters in relation to road conditions and prior riding experience","authors":"Hyunchul Park, Taeho Oh, Jaehyuck Lim, Inhi Kim","doi":"10.1016/j.trf.2025.103482","DOIUrl":"10.1016/j.trf.2025.103482","url":null,"abstract":"<div><div>While micromobility offers flexible solutions for first and last mile transportation, harmonizing with the existing public transportation, it raises substantial safety concerns, especially when overtaken by vehicles. Since the road infrastructure and regulations for micromobility have not kept pace with the rapid expansion, understanding the unique safety issues related to micromobility remains a critical task. This study aims to investigate the influence of cross-modal experience—drivers' direct experience with micromobility and road conditions on overtaking behavior towards micromobility, which includes e-scooters and bicycles, representatively. Using a VR-based simulator experiment, data were collected from 30 participants to analyze overtaking behavior. A linear mixed model was applied to assess how these factors impact driver's behaviors, such as lateral clearance, overtaking speed, and gaze fixation counts. The results indicate that drivers with cross-modal experience maintain wider lateral clearances when overtaking both e-scooters and bicycles, reflecting heightened awareness of micromobility safety needs. Additionally, poor road conditions make drivers increase lateral clearance and gaze fixation, emphasizing the importance of road infrastructure in ensuring safety. However, cross-modal experience did not significantly affect overtaking speed and road conditions for e-scooters. Conversely, a significant interaction with bicycles highlights the complexity of driving behavior and the need for comprehensive safety strategies. These findings support the development of educational programs promoting cross-modal experiences and targeted infrastructure improvements, such as dedicated e-scooter lanes, to enhance road safety for all users.</div></div>","PeriodicalId":48355,"journal":{"name":"Transportation Research Part F-Traffic Psychology and Behaviour","volume":"117 ","pages":"Article 103482"},"PeriodicalIF":4.4,"publicationDate":"2025-12-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145796856","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Driving under the influence of cannabis (DUIC) is a growing public health concern, particularly among young drivers. This pilot study explores the short-term preliminary efficacy of High Alert, a brief smartphone intervention designed to reduce DUIC among youth.
Methods
An online pilot randomized controlled trial was conducted with 102 youth aged 18–24 who had a history of DUIC (≥3 times in the past 3 months). Participants were randomized into three groups: High Alert (n = 37), Active Control (n = 34), or Passive Control (n = 31). High Alert included two web-based sessions on cannabis and DUIC education. The Active Control received a single session reviewing six DUIC-related infographics, while the Passive Control received no intervention. The primary outcome was self-reported DUIC incidents (alone or with other substances) over 3 months, assessed at baseline and 3-month follow-up.
Results
Among the 52 participants who completed the 3-month follow-up (High Alert: n = 16; Active Control: n = 16; Passive Control: n = 20), High Alert showed the greatest mean reduction in DUIC incidents (-7.44, Cohen’s d = -0.40), compared to Active Control (-3.62, d = -0.49) and Passive Control (-3.05, d = -0.38). The reduction was statistically significant compared to Passive Control (β = -0.61, p = .03), but not Active Control (β = -0.08, p = .781).
Conclusions
Preliminary findings suggest that High Alert may show promise in reducing self-reported DUIC behaviours compared to a no-contact control group, but additional research with larger samples and longer follow-ups is needed.
{"title":"Preliminary efficacy of ‘high-alert’ a brief smartphone intervention to reduce Cannabis-impaired driving among youth: A pilot randomized controlled trial","authors":"Robert Colonna , Patricia Tucker , Angela Mandich , Liliana Alvarez","doi":"10.1016/j.trf.2025.103489","DOIUrl":"10.1016/j.trf.2025.103489","url":null,"abstract":"<div><h3>Introduction</h3><div>Driving under the influence of cannabis (DUIC) is a growing public health concern, particularly among young drivers. This pilot study explores the short-term preliminary efficacy of High Alert, a brief smartphone intervention designed to reduce DUIC among youth.</div></div><div><h3>Methods</h3><div>An online pilot randomized controlled trial was conducted with 102 youth aged 18–24 who had a history of DUIC (≥3 times in the past 3 months). Participants were randomized into three groups: High Alert (n = 37), Active Control (n = 34), or Passive Control (n = 31). High Alert included two web-based sessions on cannabis and DUIC education. The Active Control received a single session reviewing six DUIC-related infographics, while the Passive Control received no intervention. The primary outcome was self-reported DUIC incidents (alone or with other substances) over 3 months, assessed at baseline and 3-month follow-up.</div></div><div><h3>Results</h3><div>Among the 52 participants who completed the 3-month follow-up (High Alert: n = 16; Active Control: n = 16; Passive Control: n = 20), High Alert showed the greatest mean reduction in DUIC incidents (-7.44, Cohen’s d = -0.40), compared to Active Control (-3.62, d = -0.49) and Passive Control (-3.05, d = -0.38). The reduction was statistically significant compared to Passive Control (β = -0.61, p = .03), but not Active Control (β = -0.08, p = .781).</div></div><div><h3>Conclusions</h3><div>Preliminary findings suggest that High Alert may show promise in reducing self-reported DUIC behaviours compared to a no-contact control group, but additional research with larger samples and longer follow-ups is needed.</div></div><div><h3><span><span>ClinicalTrials.gov</span><svg><path></path></svg></span> registration</h3><div><span><span>NCT06098573</span><svg><path></path></svg></span>.</div></div>","PeriodicalId":48355,"journal":{"name":"Transportation Research Part F-Traffic Psychology and Behaviour","volume":"117 ","pages":"Article 103489"},"PeriodicalIF":4.4,"publicationDate":"2025-12-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145796858","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Mobile phone use while riding is a common risky behavior among Iranian delivery riders, posing a significant risk to their safety. This study examines delivery riders' intentions and self-reported behavior regarding using work-related apps, calling, and using navigation apps based on an extended Theory of Planned Behavior (TPB) model. The extended model considered riding in extreme conditions, mobile phone involvement, risk perception, and distraction perception in explaining mobile phone use. To gain a deeper understanding of the salient beliefs, open-ended questions were also provided. Data gathered from 419 delivery riders using face-to-face survey interviewing in Tehran, Iran and were examined using Structural Equation Modeling (SEM). Additionally, Multiple Indicator Multiple Causes (MIMIC) models and Multi-Group Analysis (MGA) were employed to address heterogeneity across various individual groups. The results revealed that behavioral intention, followed by riding in extreme conditions and perceived behavioral control, were the strongest predictors of mobile phone use across all models. However, attitude was significant only for work-related apps and calling. In addition, risk perception and distraction perception indirectly influenced mobile phone use through perceived behavioral control for all three behaviors, while the effect of mobile phone involvement was strongest for calling behavior and negligible for work-related and navigation apps. The results showed that riders with prior crash experience used work-related apps and calls more, while experienced riders and those with longer working hours used navigation apps less. Based on the findings of this study, several implications, including targeted interventions and suggestions for employers to reduce mobile phone use, were proposed.
{"title":"Investigating factors influencing different types of risky mobile phone use among delivery riders: Using work-related apps, calling, and using navigation apps","authors":"Maedeh Rajabi , Zeinab Karami , Kayvan Aghabayk , Sina Rejali , Massoud Palassi","doi":"10.1016/j.trf.2025.103490","DOIUrl":"10.1016/j.trf.2025.103490","url":null,"abstract":"<div><div>Mobile phone use while riding is a common risky behavior among Iranian delivery riders, posing a significant risk to their safety. This study examines delivery riders' intentions and self-reported behavior regarding using work-related apps, calling, and using navigation apps based on an extended Theory of Planned Behavior (TPB) model. The extended model considered riding in extreme conditions, mobile phone involvement, risk perception, and distraction perception in explaining mobile phone use. To gain a deeper understanding of the salient beliefs, open-ended questions were also provided. Data gathered from 419 delivery riders using face-to-face survey interviewing in Tehran, Iran and were examined using Structural Equation Modeling (SEM). Additionally, Multiple Indicator Multiple Causes (MIMIC) models and Multi-Group Analysis (MGA) were employed to address heterogeneity across various individual groups. The results revealed that behavioral intention, followed by riding in extreme conditions and perceived behavioral control, were the strongest predictors of mobile phone use across all models. However, attitude was significant only for work-related apps and calling. In addition, risk perception and distraction perception indirectly influenced mobile phone use through perceived behavioral control for all three behaviors, while the effect of mobile phone involvement was strongest for calling behavior and negligible for work-related and navigation apps. The results showed that riders with prior crash experience used work-related apps and calls more, while experienced riders and those with longer working hours used navigation apps less. Based on the findings of this study, several implications, including targeted interventions and suggestions for employers to reduce mobile phone use, were proposed.</div></div>","PeriodicalId":48355,"journal":{"name":"Transportation Research Part F-Traffic Psychology and Behaviour","volume":"117 ","pages":"Article 103490"},"PeriodicalIF":4.4,"publicationDate":"2025-12-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145796915","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Motorcycle taxi riders (MTRs) are vital for urban mobility in many low- and middle-income countries (LMICs), yet they experience a disproportionately high rate of involvement in road traffic crashes. This is primarily linked to risky riding behaviors whose underlying drivers remain insufficiently understood.
Objective
This study examines the prevalence of risky riding behaviors among MTRs in Dar es Salaam (DSM), Tanzania, and identifies the key motivations driving these behaviors.
Methods
A cross-sectional survey was conducted with 853 MTRs in urban and peri-urban areas of DSM. Data were analyzed using descriptive statistics, Spearman correlation, and logistic regression to explore associations between reported behaviors and motivational factors.
Results
Key predictors of risky behaviors included financial pressure, passenger and peer influence, and deficits in risk perception. Riders under financial strain were more likely to overload passengers, ignore traffic signals, and speed. Influences from peers and passengers also led to behaviors such as lane splitting, riding on restricted roads, and using phones while riding. Despite weak enforcement, some riders demonstrate a degree of self-regulation. However, training programs and public awareness campaigns did not consistently lower risk levels.
Conclusion
Risky behaviors among MTRs in DSM are often rational responses to local economic and systemic constraints rather than just recklessness. Effective interventions must move beyond punitive measures to encompass training reform, infrastructure improvements, inclusive enforcement, and financial support. To enhance safety, it is essential to engage riders as partners in road safety rather than merely subjects of regulation.
{"title":"Determinants of risky riding behaviors among motorcycle taxi riders in Dar Es Salaam, Tanzania","authors":"Katondo Salvatory Nambiza , An Neven , Wilfred Gordian Kazaura , Kris Brijs","doi":"10.1016/j.trf.2025.103486","DOIUrl":"10.1016/j.trf.2025.103486","url":null,"abstract":"<div><h3>Background</h3><div>Motorcycle taxi riders (MTRs) are vital for urban mobility in many low- and middle-income countries (LMICs), yet they experience a disproportionately high rate of involvement in road traffic crashes. This is primarily linked to risky riding behaviors whose underlying drivers remain insufficiently understood.</div></div><div><h3>Objective</h3><div>This study examines the prevalence of risky riding behaviors among MTRs in Dar es Salaam (DSM), Tanzania, and identifies the key motivations driving these behaviors.</div></div><div><h3>Methods</h3><div>A cross-sectional survey was conducted with 853 MTRs in urban and peri-urban areas of DSM. Data were analyzed using descriptive statistics, Spearman correlation, and logistic regression to explore associations between reported behaviors and motivational factors.</div></div><div><h3>Results</h3><div>Key predictors of risky behaviors included financial pressure, passenger and peer influence, and deficits in risk perception. Riders under financial strain were more likely to overload passengers, ignore traffic signals, and speed. Influences from peers and passengers also led to behaviors such as lane splitting, riding on restricted roads, and using phones while riding. Despite weak enforcement, some riders demonstrate a degree of self-regulation. However, training programs and public awareness campaigns did not consistently lower risk levels.</div></div><div><h3>Conclusion</h3><div>Risky behaviors among MTRs in DSM are often rational responses to local economic and systemic constraints rather than just recklessness. Effective interventions must move beyond punitive measures to encompass training reform, infrastructure improvements, inclusive enforcement, and financial support. To enhance safety, it is essential to engage riders as partners in road safety rather than merely subjects of regulation.</div></div>","PeriodicalId":48355,"journal":{"name":"Transportation Research Part F-Traffic Psychology and Behaviour","volume":"117 ","pages":"Article 103486"},"PeriodicalIF":4.4,"publicationDate":"2025-12-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145796957","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号