This paper examines the potential effect of cell phones on motor vehicle fatality rates normalized for other driving related and socioeconomic factors. The model used is non-linear so as to address both life-taking and life-saving attributes of cell phones. The model is evaluated using classical methods along with Bayesian Extreme Bounds Analysis (EBA). The use of both classical and Bayesian methods diminishes the model and parameter uncertainties which afflict more conventional modeling methods which rely on only one of the two methods. The results indicate the presence of both life-taking and life-saving attributes of cell phones on motor vehicle fatality rates depending on the volume of cell phone subscribers in existence. (Author/publisher)
Abstract