Fatal distracted driving pedestrian-involved crashes in Ghana: Exploring cluster-specific factor associations using cluster correspondence analysis

Distracted driving presents a serious threat to pedestrian safety, especially in low- and middle-income countries (LMICs) where pedestrian infrastructure is limited. Although distraction-related vehicle-pedestrian crashes result in numerous fatalities each year, few studies have examined the factors contributing to these incidents, particularly in LMIC contexts. This study addresses this research gap by exploring the latent patterns associated with fatal pedestrian crashes caused by distracted driving in Ghana. A robust unsupervised machine learning method, Cluster Correspondence Analysis (CCA), was used to analyze 1638 fatal crash records from 2014 to 2018. This method simultaneously classifies crash observations into homogeneous clusters and identifies key factor associations within each cluster. The analysis revealed seven distinct clusters based on pedestrian manoeuvres, crash location, and vehicle type. Findings indicate that most fatal distraction-related pedestrian crashes occurred during crossing manoeuvres, often involving privately operated vehicles in daylight and clear weather. One cluster revealed that older heavy vehicle drivers (aged 45–64) were often involved in crashes with younger pedestrians (under 25) walking along the road. Another pattern showed that female drivers were mainly involved in fatal multi-vehicle crashes at night, typically during turning manoeuvres at intersections with medians. Additionally, crashes involving buses and minibuses were frequently associated with brake failure, involvement of multiple vehicles, older drivers (65+), and a higher likelihood of being hit-and-run incidents. This study contributes to pedestrian safety research by uncovering crash-specific patterns associated with distracted driving in an LMIC setting. Insights obtained from the study are discussed, and targeted education, engineering, encouragement, and enforcement countermeasures are developed based on the study's findings. • Cluster-specific factors linked with fatal distracted-driving pedestrian crashes are explored. • Pedestrian crossing crashes caused by distracted car drivers offen occur during the day. • Distracted taxi drivers mostly hit crossing pedestrians at night during bad weather. • Fatal crashes caused by trucks are likely to involve pedestrians walking along the road. • Elderly bus drivers are associated with hit-and-run multi-casualty crashes at night.