Existing trust management schemes for Vehicular Ad Hoc Networks (VANETs) can be classified into four main categories: role based, experience based, priority based, and majority opinion. To our knowledge, there is no study that compares the reliability of each of the four trust models to accurately report an accident. In this paper, we perform a comparative simulation study to measure the accuracy of each trust model to report an accident in a road. We simulate road accidents with varying vehicle densities using Manhattan mobility model. We report on how to measure the relative accuracy of four existing trust management schemes for VANETs. Our simulation results show that experience based trust outperforms other trust models in accurately reporting accidents in a road. We also demonstrate that for Majority Based Trust, setting the threshold at 20% enables more vehicles to reliably learn about an accident than when a 50% threshold is used.