Development of a hazard prediction test for Chinese cyclists and its association with crash involvement.

INTRODUCTION

Hazard perception is the ability to sense the road environment and predict and respond to the upcoming dangerous events, and this factor is closely related to cyclists' crash involvement. However, due to a lack of valid instruments, studies concerning the hazard predictions and crashes of cyclists in China remain limited. This study attempted to develop a hazard prediction test for cyclists.

METHOD

The experiment presented 44 video clips filmed from cyclists' perspectives to 61 children aged 13-16 years and 119 adults aged 18-30 years. After the video clip displayed a black screen, participants were asked to answer the following questions: "What is the hazard?" (question 1), "Where is the hazard?" (question 2), and "What happens next?" (question 3). The differences in test scores between cyclists with high and low levels of experience, cyclists of different ages, and crash-involved and noncrash-involved cyclists were compared to examine the validity of the test.

RESULTS

The final test contained 21 video clips and the internal consistency reliabilities of the three questions were satisfactory. Experienced cyclists are better able to identify and predict hazards than are cyclists with less experience, and adult cyclists have better hazard prediction abilities than child cyclists. More importantly, crash-involved cyclists receiving lower scores with respect to their ability to identify and predict hazards than noncrash-involved cyclists, thus indicating that the discriminant validity of the test was acceptable.

PRACTICAL APPLICATIONS

This newly-developed test exhibits acceptable reliability and validity and can be used as an effective tool to measure the hazard prediction abilities of cyclists with different ages and levels of experience in China. The natural driving videos and test in the study may hold promise for future practical applications, and the implications for road safety are discussed.