Statistical models for count data with applications to road accidents in Ghana
Loading...
Date
June, 2015
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Road accidents in Ghana seems to be on ascendency and the root causes of these
accidents have been attributed to issues such as human errors and superstitions.
Since the occurrence of accidents are discrete, they are often modeled using
count regression models. It is therefore the purpose of this study to determine
an appropriate count regression model that adequately fits road accidents on
urban roads in Ghana and to determine the key predictors of road accidents
using the appropriate count model with respect to the expected number of
person killed in an accident. Several models fitted using count data (occurrences
of road accidents) in the field of transportation were compared.These models
include Poisson,Negative Binomial and Conway-Maxwell-Poisson count regression
models . To compare the performance of these models, the various model
selection methods such as Deviance goodness of fit, Akaike’s Information Criterion
(AIC)and Bayesian Information Criterion (BIC) were employed. Because the
values of the Deviance goodness of fit, AIC and BIC respectively of the Negative
Binomial was the smallest as compared to that of Conway-Maxwell-Poisson and
Poisson models, it appeared that the Negative Binomial model performed best
as compared to the Poisson and the CMP model. Base on the appropriate
count regression model selected (Negative Binomial model) the key predictors
that contributed significantly and had a high effect on the expected number of
persons to be killed in a road accidents within a particular time were Head-on
collision as Collision type, Improper-overtaking and Loss of control as Driver
errors, Bus/minibus as Type of vehicle, Fig/midst as Weather condition and
Night with street lights off as Light condition.
Description
A Thesis submitted to the Department of Mathematics,
Kwame Nkrumah University of Science and Technology In
partial fulfillment of the requirement for the Degree of M.Phil Mathematical Statistics, 2015