Data are provided by the Western Pennsylvania Data Regional Center (WPRDC). Beaver County Crash data is just one of the thousands of datasets available in the website.
This webpage has been created as short byte of the 05-839 class at CMU in Spring'17. All the work showed here is based on the tutorial designed by Jen Mankoff and Nikola Banovic.
If you want to work directly with the whole dataset, just visit the Google Fusion Table and create your own visualization and story.
People can assume more accident happen on road sections with the highest AADT, but it is not the case. Explore the map to find inconsistences on that assumption. You will discover particular sections with a high concentration of accidents.
Trips patterns and origin-destination pairs changes depending on day of the week. The usual weekday pattern is modified. Check how around golf clubs there are only accidents during the weekends.
The distribution of accidents is pretty uniform. Considering that the level of traffic over the weekends is lower, there is a higher level of accident over the weekends when comparing to the AADT data.
Traffic accidents hide not only a spatial meaning considering where they happen over the time, but as well, segmentation by age can tell lots of things about the lyfestyles of people and the risky ranges. It seems that 16 years old drivers use mostly their car for going out on weekends, and not for going to the work or to the high school as older teenagers. However, when people turn 19 years old, again there is an intense use of cars over the weekends. Due to the high level of crashes, we can assume that somehow it is related to going out on Fridays night.
By studying and analyzing the location of youngest drivers accidents we can detect that actually they are not in the more dangerous sections, where more accidents happen, but on other locations. It would be interesting to cross those locations with the locations of institutes.