Predicting Vehicle Occupancies from Accident Data: An Accident Severity Approach
- 1 January 1998
- journal article
- Published by SAGE Publications in Transportation Research Record: Journal of the Transportation Research Board
- Vol. 1635 (1) , 93-104
Abstract
Past studies have shown that using accident records to estimate vehicle occupancies (i.e., using the observed occupancies of vehicles involved in accidents) results in an overestimation of occupancy. There are a number of possible reasons for this, one of which is that multioccupant accidents are more likely to be reported (i.e., appear in an accident database) because, with more people, the possibility of an injury is greater. The interaction between vehicle occupancy and accident severity is used to develop a method to correct for the occupancy overestimation bias inherent in accident data. A nested logit model of occupancy and severity is estimated, and a correction technique is applied to eliminate biases. The results show that the proposed approach gives accurate predictions of vehicle occupancies using standard accident data.Keywords
This publication has 5 references indexed in Scilit:
- Predicting the severity of motor vehicle accident injuries using models of ordered multiple choiceAccident Analysis & Prevention, 1996
- An exploratory multinomial logit analysis of single-vehicle motorcycle accident severityJournal of Safety Research, 1996
- Statistical analysis of accident severity on rural freewaysAccident Analysis & Prevention, 1996
- Risk factors of fatality in motor vehicle traffic accidentsAccident Analysis & Prevention, 1994
- Predicting injury risk with “New Car Assessment Program” crashworthiness ratingsAccident Analysis & Prevention, 1988