Modeling Vehicle-pedestrian Crashes With Excess Zero Along Malaysia Federal Roads

• Published in: Procedia, social and behavioral sciences Vol. 53; pp. 1216 - 1225
• Main Authors: Pour, Mehdi Hossein, Prasetijo, Joewono, Yahaya, Ahmad Shukri, Ghadiri, Seyed Mohammad Reza
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 03.10.2012
• Subjects: crash prediction models; model selection criteria; two-part models; Vehicle-pedestrian crashes; Vehicle-pedestrian crashes; crash prediction models; model selection criteria; two-part models
• ISSN: 1877-0428; 1877-0428
• DOI: 10.1016/j.sbspro.2012.09.970

This study focused on pedestrian crashes occurred in a 543-km sample of Malaysia federal roads in which roadways mostly pass through rural areas. To do this, four count models including Poisson, negative binomial (NB), zero-inflated Poisson (ZIP), and zero-inflated negative binomial (ZINB) models were developed to establish the relationship between vehicle-pedestrian crashes and the relevant road infrastructure and environmental features. Data on pedestrian crashes were collected for a four-year period from 2007 through 2010. The result indicates that ZIP model performed best, in terms of the comparative criteria.