Understanding the factors affecting the frequency and severity of aging population-involved crashes in Florida

External sources such as traffic flow and light conditions and weather-related environmental factors are known to have significant effects on crash occurrences of aging drivers. While most studies in this area considered the effect of only the Average Annual Daily Traffic, traffic flow may vary over time and this variation may have a significant impact, especially on vulnerable populations such as aging people. In this paper, we investigate a logistic regression crash prediction model that quantifies the impact of hourly traffic flow on the aging-involved crashes with respect to different weather and light conditions. For training the model, we propose to incorporate several independent data sets, including meteorological and hourly traffic flow-related data with a specific focus on the differences between aging population-involved crashes and those involving other age groups. Logistic regression models at various geographical locations are fit to these data sets in order to investigate the effect of the environmental conditions and traffic flow on the frequency and severity of aging-involved crashes. Adequacy and predictive capability of the fitted models are demonstrated with goodness of fit tests and cross-validation studies.