In this paper, we present a statistical analysis based on the 2k factorial methodology to determine the representative factors affecting traffic safety applications in Vehicular ad hoc networks (VANETs). Our purpose is to determine what are the key factors affecting Warning Message Dissemination (WMD) in order to concentrate on such parameters, thus reducing the amount of required simulation time when evaluating VANETs. Simulation results show that the key factors affecting warning messages delivery are: (i) the transmission range, (ii) the radio propagation model used, and (iii) the density of vehicles. Based on this statistical analysis, we evaluate a compound key factor: neighbor density. This factor combines the above-mentioned factors into a single entity, reducing the number of factors that must be taken into account for VANET researchers to evaluate the benefits of their proposals.