Using logistic regression and point-biserial correlation, an investigation of pedestrian violations and their opportunities to cross at signalized intersections

Abstract

Intersection's safety and efficiency are affected by pedestrian offences. As a result, it's critical to recognize and prioritize the factors that influence pedestrian violations. Three intersections with commercial, medical, and educational land-uses, as well as a large volume and variety of pedestrians, were chosen for this paper. The main goal is to determine priorities among the factors under consideration in order to make a more cost-effective suggestion in the executive field, and second, to determine how much the ability to cross at the time of pedestrian arrival influences his/her decision, along with the other influential factors. Videotaping was used to collect factors such as pedestrian characteristics, traffic signals, intersection geometry, various peak hours of working days, and weather conditions. Data collecting is quantitative, primary, and descriptive. Statistical analysis methods are used to test relationships between variables. The decision-making process of pedestrians is modelled using logistic regression. The findings reveal that the time of arrival, median type, and time of day have a significant effect on pedestrians' likelihood of violation. It can also be seen that pedestrians' opportunities to cross (smaller gaps between vehicles) are associated with higher compliance probabilities. An effective median design could have a significant impact on pedestrians' willingness to wait for a safe crossing during the red phase based on field observations.