Effect of Different Types of Distractions on Driving Behavior and Road Safety Using Naturalistic Driving Data Under Mixed Traffic Conditions

Distracted driving and excessive speeding are among the major contributors to traffic accidents and fatalities worldwide. This study assesses how drivers regulate their speed under mixed traffic conditions (MTC) while engaging in distracted driving behaviour (DDB). Driver distractions such as visual, manual and cognitive arise from engagements like mobile phone use, eating, drinking, eyes off the road and other DDB. Using naturalistic driving data collected from 41 drivers, this study investigates the relationship between involvement in DDB and speed selection in MTC. Linear mixed-effects (LME) models were used to analyse speed-related measures while accounting for repeated observations within drivers and inter-driver variability. The models amalgamated different types of distractions, their duration and driver attributes such as age, education level and driving experience, along with relevant interaction effects. The final models identified significant interactions between distraction types and driving attributes, underscoring the heterogeneous nature of distracted driving behaviour. Statistical analysis using paired t-tests and linear mixed effect models revealed significant differences between distracted and undistracted driving conditions. Distracted driving was associated with reduced acceleration and deceleration activity, indicating lower control intensity. The median speed model showed that cognitive distractions, such as talking to oneself or singing, and increased distraction duration are associated with higher driving speeds. In contrast, the speed variability model indicated that visual distractions, such as looking externally or internally, are associated with reduced speed fluctuations during short-duration events but increased variability with prolonged distraction, reflecting a decline in driving stability. Interaction effects further highlighted that the impact of distraction varies across driver characteristics and conditions. Overall, the findings suggest that distraction influences both speed magnitude and control, with prolonged distraction leading to reduced stability and increased safety risk.