Citation

Jiang X, Lyles RW. Accid. Anal. Prev. 2007; 39(4): 649-656.

Affiliation

Department of Civil and Environmental Engineering, Michigan State University, East Lansing, MI 48824-1226, Unites States.

Copyright

(Copyright © 2007, Elsevier Publishing)

DOI

10.1016/j.aap.2006.10.011

PMID

17118322

Abstract

The attractiveness of quasi-induced exposure lies in the simplistic nature of its theory and application. As opposed to vehicle miles of travel, quasi-induced exposure is developed solely from the accident data themselves. Involvement ratios (IRs) are used to describe the relative over- or under-involvement of different driver-vehicle combinations in traffic accidents. The issue of systematic bias in the involvement ratios is explored, and it is shown both theoretically and empirically that the variation in average speeds between vehicle types can affect relative accident involvement ratios. For vehicle types that routinely travel faster, the IRs will likely be underestimated; while for slower-moving vehicle types, the IRs will be overestimated. In the case of speed, the magnitude of the effect on the IRs is dependent both on the magnitude of the speed differential and the percentage of slower vehicles in the traffic stream. Conclusions can be extended to whenever there are speed or other behavior disparities associated with the driver-vehicle combinations in the traffic stream. Other examples include the speed discrepancy caused by different drivers (e.g., younger versus older drivers). Caution must be used in interpreting results from applications of the quasi-induced technique whenever such biases might be encountered.

Language: en