Topic Modeling Analysis of Driving Fears: Individual Variations and Emotional Responses

Driving fears can significantly impact emotional responses and behavior on the road, influencing both individual performance and overall traffic safety. In this study, 202 participants were presented with 64 vignettes of diverse driving conditions, and asked how driving in similar conditions would make them feel. Forty three participants reported and explained having specific driving fears. Latent Dirichlet Allocation (LDA) was applied to these responses and four key themes were identified: post-accident anxiety, fear of driving on the highway and merging onto the highway, truck-related and nighttime driving fears, and fear of driving in adverse weather. On average, participants who reported having fear from driving in adverse weather conditions, reported lower valence and arousal to the vignettes representing similar conditions. Furthermore, individual differences significantly influenced emotional responses across different topics. For example, elderly participants exhibited lower valence but higher arousal than younger participants across most fear topics. Additionally, those who often had companions in the car and feared driving in adverse weather reported lower arousal and more valence than those who often drove alone. These findings highlight the importance of personalized interventions, such as driver-assist systems designed to support drivers in managing specific fears and enhancing their confidence in maintaining control, which ultimately improves overall road safety.