Identifying and addressing multi-source database inconsistencies: Evidences from global road safety information

Abstract

The data collection and the creation of extensive databases for the investigation of different global phenomenon, such as road traffic fatalities, inherent risks of information inconsistencies. The current paper is presenting a novel approach on the efficiently detection of potential data anomalies (inconsistencies), using wide in range socio-economic factors from different ‘instances’ years (2010 and 2013), for the investigation of the phenomenon of road traffic fatalities concerning 121 UN countries (restricted to UN countries with significant population). Unfortunately, collecting information from different, even reliable, sources (global organizations) raises speculations of uncertain, implausible, inconsistent and unstable information, which can be transparent with different data-model analysis likewise Principal Component Analysis, Negative Binomial regression analysis and Structural Equation Modeling.