The analysis does not consider all the dataset's present assumptions, creating a bias in the results. For example, the roadway ratings are subjective, and there are no universal requirements for the measurement. The number of crashes that occurred within the intersected poor-rated roadway segments appears to have a larger number compared to the total number of fatalities and injuries; however, the NYPD raw data might include some errors, which the analysis did not account for. For example, collision data might have been recorded twice with a different point location because it is recorded at borough boundaries, resulting in counting the crash twice. There are other contributing factors that both data sets do not account for, such as the roadway/intersection geometry, traffic, and pedestrian volume. The dataset lacked some of the driver's information and the vehicle characteristic, and street conditions. With additional information, a more precise, detailed analysis could be carried out in the future.