top of page
Search

(Python) Crash Into Me: 2016-2020 Vehicle Crash (Exploratory) Analysis

Updated: May 27, 2022

Catherine Greene

April 2022


Accidents, or unintentional injuries, are in the top 10 leading causes of death in the United States. The term accidents is quite broad but in this instance it includes, but is not limited to, transport accidents such as motor vehicle accidents, non-transport accidents such as falls, intentional self-harm, and assault. This analysis will be exploring 2016-2020 vehicle accident reports to answer some key questions.



The goal of this analysis is to answer:

  • Have fatal vehicle crashes been increasing or decreasing in past years?

  • What car makes and model combinations were involved in the most accidents?

  • What manufacturer is involved in the most accidents?

  • Are there different states where the most fatal vehicles are more accident prone?


The following databases will be used in the analysis:

  • accident_20.csv

  • vPicDecode.csv for years between 2016 and 2020

  • Fatality Rates.csv

These are raw datasets retrieved from the National Highway Traffic Safety Administration (NHTSA).




Cleaning the Data

Fatality counts from 2020 were included in the Fatality Rates table of 1994 through 2019 data. This will allow us to view the increase or decrease in fatalities. The seperate vPicDecode.csv dataframes for the years between 2016 and 2020 have been imported and modified to include the manufacturer, make, and model involved in each accident.


Final Datasets (Snippets)

Following are the final cleaned dataferames that will be used in the analysis.












Exploratory Data Analysis

While remaining below or near the average number of accident fatalities for the years between 1994 and 2020, the number of fatal accidents occurring in the United States have been steadily increasing since 2015. Despite Covid-19, the most fatal year between 2016 and 2020 was in fact 2020 with 38,824 fatalities. This was a 2,728 increase from 2019, a 6,345 increase from the lowest fatality year (2011), and 23 fatalities above the mean (38,801 fatalities).


Between 2016 and 2020, fatalities have all remained within 2,730 fatalities from previous and consecutive years. As shown in the above graph, there was quite a dip in 2019 (36096 fatalities) and during the pandemic there was a significant rise. While many factors attribute to car accidents and fatalities, this analysis will now explore the different car makes, models, and manufacturers that were involved.



The above graph shows the number of accidents between the years 2016 and 2020 for the top four car models. The car with the most accidents was the Chevrolet Silverado followed by the Ford F-150. These cars were consistently in the top four for accidents when the years were looked at separately. The manufacturers for the Chevrolet Silverado, Ford F-150, Honda Accord, and Toyota Camry are General Motors, Ford, Honda, and Toyota respectively.



The top five manufacturers involved in accidents between 2016 and 2020, were General Motors, Ford, FCA, Toyota, and Honda which is somewhat unsurprising as these manufacturers, aside from FCA, hold the spots for the top four car makes with accidents. The percentages in the above graph, include all car makes, not just the top five meaning 21.37% of all accidents in 2016- 2020 involved a General Motors vehicle.


The next step in this analysis was to explore if there were states where the top four vehicles for accidents were more accident prone.




The states with the most accidents for the top cars were consistently Texas, California, and Florida which translates over to the graph on the right that shows the states with the most accidents between 2016 and 2020 for all vehicles. Texas with 26,168 and California with 26,050 accidents were neck and neck and Florida with 22,712 accidents was a close third.


Conclusion

This data was analyzed in order to gain insight into one of the top 10 leading causes of death: accidents. As seen in the beginning stages of the exploratory data analysis, fatal accidents, though ultimately lower than the average since 1994, have started to increase in the last five years. This led to a basic exploration of the vehicles involved in these crashes.

The cars with the higher counts of accidents are consistently the Chevrolet Silverado, Ford F-150, Honda Accord, and the Toyota Camry. When looking at the number of accidents in each state, with involvement of the previously mentioned cars, it is clear that they mostly occur in Texas, California, and Florida. The top manufacturers involved in the most accidents are General Motors and Ford.



Next Steps

While answering questions regarding the initial goals, this analysis simultaneously posed further questions. Knowing that cars alone do not cause accidents, does this analysis show the cars that are more likely to be in accidents or does it shed light on the most popular cars in the United States? Similarily, it is very possible that there is less of a causation between manufacturers and accidents and more of a correlation between more popular manufacturers or number of cars on the road by a particular manufacturer and the number of accidents involving said manufacturer. Are there specific types of vehicle crashes that occur more often in the top accident states? How many of the vehicles most often involved in accidents are sold yearly? Is there a correlation between vehicle sales and number of accidents?

Just some of the great questions worth looking into in a future analysis.



Citation

NHTSA file downloads. NHTSA. (n.d.). Retrieved April 6, 2022, from https://www.nhtsa.gov/file-downloads?p=nhtsa/downloads/FARS/2020/ NHTSA file downloads. NHTSA. (n.d.). Retrieved April 6, 2022, from https://www.nhtsa.gov/file-downloads?p=nhtsa/downloads/FARS/2019/ NHTSA file downloads. NHTSA. (n.d.). Retrieved April 6, 2022, from https://www.nhtsa.gov/file-downloads?p=nhtsa/downloads/FARS/2018/ NHTSA file downloads. NHTSA. (n.d.). Retrieved April 6, 2022, from https://www.nhtsa.gov/file-downloads?p=nhtsa/downloads/FARS/2017/ NHTSA file downloads. NHTSA. (n.d.). Retrieved April 6, 2022, from https://www.nhtsa.gov/file-downloads?p=nhtsa/downloads/FARS/2016/ Administration, N. H. T. S. (n.d.). Fars Encyclopedia. FARS Encyclopedia: Trends - General. Retrieved April 6, 2022, from https://www-fars.nhtsa.dot.gov/Trends/TrendsGeneral.aspx




 
 
 

Comments

bottom of page