Average vehicle occupancy (AVO) is a major measure used to assess the effectiveness of transportation demand management (TDM) techniques and programs. Traditionally, vehicle occupancy data are collected using the field data collection methods such as the roadside windshield method and the carousel method. While these field collection methods have the benefit of being more up to date and can be tailored to specific application needs, they are also much more costly to conduct. This paper describes a study that applies the passenger information in accident records to analyze the averagevehicle occupancy trends in Florida as they relate to different years, months, days of week, hours, driverÆs age group, driverÆs gender, driverÆs race, urbanization, roadway functional class, and weather conditions. Bothparametric and non-parametric analyses of variance (ANOVA) were applied and it was found that AVOs were affected by all of the factors considered to various degrees. In terms of temporal factors, the AVOs in Florida was found to decline steadily over a 15-year period examined, AM peaks experienced lower AVOs than PM peaks, and weekday AVOs were significantly lower than weekend AVOs. In terms of socioeconomic factors, younger drivers werefound to have higher AVOs than older drivers, female drivers tend to havemore passengers in their vehicles than male drivers, and Blacks and Hispanic Americans tend to have higher AVOs than White Americans. Other findings include higher AVOs in rural areas than urban areas, higher AVOs under poor weather conditions, and higher AVOs for rural freeways than urban freeways.
Abstract