Title

Nature of Modeling Boundary Pedestrian Crashes at Zones

Authors

Authors

C. Siddiqui;M. Abdel-Aty

Comments

Authors: contact us about adding a copy of your work at STARS@ucf.edu

Abbreviated Journal Title

Transp. Res. Record

Keywords

COLLISION PREDICTION MODELS; SPATIAL-ANALYSIS; INJURY COLLISIONS; LAND-USE; SAFETY; TRANSFERABILITY; Engineering, Civil; Transportation; Transportation Science & Technology

Abstract

Traffic analysis zones are often delineated by the existing street network. This practice may result in a considerable number of crashes on or near zonal boundaries. Although the traditional macrolevel approach to crash modeling assigns zonal attributes to all crashes that occur within the zonal boundary, this paper acknowledges the inaccuracy resulting from relating crashes on or near the boundary of the zone to merely the attributes of that zone. This paper proposes a novel approach to account for the spatial influence of neighboring zones on crashes that occur specifically on or near the zonal boundaries. Predictive models for pedestrian crashes per zone were developed with a hierarchical Bayesian framework and with separate predictor sets for boundary and interior (nonboundary) crashes. The hierarchical Bayesian model that accounted for spatial autocorrelation was found to have better goodness-of-fit measures than did models that had no specific consideration for crashes located on or near the boundaries. In addition, the models were able to capture some unique predictors associated explicitly with interior and boundary-related crashes. For example, two variables, total roadway length with a posted speed of 35 mph and long-term parking cost, were not statistically significant from zero in the interior crash model but were significantly different from zero at the 95% level in the boundary crash model.

Journal Title

Transportation Research Record

Issue/Number

2299

Publication Date

1-1-2012

Document Type

Article

Language

English

First Page

31

Last Page

40

WOS Identifier

WOS:000313929100004

ISSN

0361-1981

Share

COinS