A Comparison of Marginal and Joint Generalized Quasi-likelihood Estimating Equations Based On the Com-Poisson GLM: Application to Car Breakdowns Data

In this paper, we apply and compare two generalized estimating equation approaches to the analysis of car breakdowns data in Mauritius. Number of breakdowns experienced by a machinery is a highly under-dispersed count random variable and its value can be attributed to the factors related to the mechanical input and output of that machinery. Analyzing such under-dispersed count observation as a function of the explanatory factors has been a challenging problem. In this paper, we aim at estimating the effects of various factors on the number of breakdowns experienced by a passenger car based on a study performed in Mauritius over a year. We remark that the number of passenger car breakdowns is highly under-dispersed. These data are therefore modelled and analyzed using Com-Poisson regression model. We use the two types of quasi-likelihood estimation approaches to estimate the parameters of the model: marginal and joint generalized quasi-likelihood estimating equation approaches. Under-dispersion parameter is estimated to be around 2.14 justifying the appropriateness of Com-Poisson distribution in modelling underdispersed count responses recorded in this study.





References:
[1] Guikema, S. " Formulating informative data -based priors for failure
probability estimation in reliability analysis", Reliability Engineering and
System Safety, Vol 92, 490-502,2007.
[2] Jowaheer,V and Mamode Khan,N. " Estimating Regression Effects in
Com-Poisson Generalized Linear Model", International Journal of Computational
and Mathematical Sciences, Vol 3:4, 2009
[3] Kadane,J , Shmueli,G, Minka,G, Borle,T and Boatwright,P. " Conjugate
analysis of the Conway Maxwell Poisson distribution", Bayesian analysis,
Vol 1, 363-374,2006.
[4] Lord,D, Guikema,S and Geedipally,S. "Application of the Conway-
Maxwell-Poisson Generalized Linear Model for Analyzing Motor Vehicle
Crashes". Accident Analysis and Prevention, Vol. 40, 1123-1134,2008
[5] Shmueli,G, Minka,T, Borle,J and Boatwright,P. " A useful distribution
for fitting discrete data, Journal of Royal Statistical Society, 2005.
[6] Road accidents in Mauritius: statistics and analysis, Ministry of Public
Infrastructure, Land Transport and Shipping Traffic management and
Road safety unit, July 2006.
[7] RAC patrol report: motoring organisation, United Kingdom, 2008