Assessing Safety Perceptions at Roundabouts in India: Using Multiple Correspondence Analysis and Regression Analysis

The potential of roundabouts to improve safety and operational efficiency has led to their widespread adoption throughout India. Even with their acknowledged advantages, roundabouts are still viewed critically by the general public, which makes it difficult to evaluate how they affect driving behavior and overall safety. The present study investigated the safety perceptions at roundabouts among its users in India, emphasizing how socio-demographic factors, vehicle types, and roundabout usage patterns influence perceived safety. A structured questionnaire survey was conducted to examine the safety perceptions of a sample of 1,250 participants by analyzing user demographics, driving experience, and varying roundabout geometric configurations. Multiple Correspondence Analysis (MCA) was employed to identify patterns among respondent categories, and an ordered probit model was developed to assess factors influencing perceived roundabout safety. Key findings revealed that male users aged 26–35 years with greater roundabout familiarity perceived roundabouts to be safer, while females, motorcyclists, and bicyclists perceived roundabouts as dangerous. Roundabout entry and complex lane configurations were perceived as hazardous, with users reporting higher risk perceptions in double- and multi-lane configurations. The ordered probit model achieved an overall prediction accuracy of 72.64%, highlighting significant relationships between demographic factors, roundabout design, and safety perception. The results of the study underscored the need for user-centered design improvements and robust safety education to foster greater roundabout usability and safety in low- and middle-income countries.