Investigating Goodness-of-Fit Test Statistics for Generalized Linear Crash Models with Low Mean Values
This study has two objectives. The first objective is to examine the accuracy and reliability of test statistics for goodness-of-fit of generalized liner crash models under low crash mean situations. The second objective intends to identify a superior test statistic for goodness-of-fit tests. The Poisson and Poisson-Gamma models are commonly used to analyze crash data. For Poisson models, this paper proposes a better test statistic that can be applied for almost all mean values, except when the mean value is extremely low, for which no test statistic can be accurate. This statistic is also rather simple and easy to use. For Poisson-Gamma models, this study finds that traditional test statistics are not accurate and robust, and a more complicated method proposed in a past study is recommended. For better illustrations, real-world data are applied to identify and analyze the performances of different test statistics.
-
Corporate Authors:
500 Fifth Street, NW
Washington, DC United States 20001 -
Authors:
- Zhang, Yunlong
- Ye, Zhirui
-
Conference:
- Transportation Research Board 86th Annual Meeting
- Location: Washington DC, United States
- Date: 2007-1-21 to 2007-1-25
- Date: 2007
Language
- English
Media Info
- Media Type: CD-ROM
- Features: Figures; References; Tables;
- Pagination: 27p
- Monograph Title: TRB 86th Annual Meeting Compendium of Papers CD-ROM
Subject/Index Terms
- TRT Terms: Crash analysis; Crash data; Crash injuries; Crashes; Distributions (Statistics); Goodness of fit; Mean (Statistics); Poisson distributions; Reliability; Statistics; Test procedures
- Uncontrolled Terms: Linear models
- Subject Areas: Data and Information Technology; Highways; Safety and Human Factors; Vehicles and Equipment; I81: Accident Statistics; I91: Vehicle Design and Safety;
Filing Info
- Accession Number: 01042553
- Record Type: Publication
- Report/Paper Numbers: 07-2640
- Files: TRIS, TRB
- Created Date: Feb 8 2007 7:20PM