Twice every year in the United States, we change our clocks by an hour. Despite being only one hour, many studies have shown that daylight saving time has more impacts on our circadian rythmn than we initally thought [2]. This change every year can affect our day to day activites such as traveling to and from work -- and most Americans travel via car (America is a car-centric country) [3]. The problem we hope to look at is how daylight saying time affect traffic accidents in America.
There have been a lot of debate about keeping daylight saving time or not, as the U.S is one of the few countries that still keep this practice. Due to this, there are a lot of data on how daylight saving time benefit and hurts us, one of which are traffic accidents. The goals of this project is to find a relationship between traffic accidents happening during daylight saving time (November to March) and traffic accidents happening during standard time (April to October). Does daylight saving time cause more traffic accident or not
This project can help promote more people to question why we even change our times and if this practice that started in WW1 is even relevant to our modern society. This project can also help contribute to saving more lives in traffic accident once a relationship between daylight saving and traffic accident is found.
The database planning to be used is this US Traffic Fatality Records data from The National Highway Taffic Safety Administrations in 2016. The factors being observed for each accident is:
First couple of rows of the dataset:
| State Name | number of motor vehicles in transport | number of person not in motor vehicles in transport | number of person in motor vehicles in transport | day of crash | month of crash | hour of crash | manner of collision | light condition |
|---|---|---|---|---|---|---|---|---|
| Alabama | 1 | 0 | 1 | 22 | 11 | 16 | Not Collision with Motor Vehicle in Transport (Not Necessarily in Transport for 2005-2009) | Daylight |
| Alabama | 1 | 0 | 2 | 30 | 3 | 14 | Not Collision with Motor Vehicle in Transport (Not Necessarily in Transport for 2005-2009) | Daylight |
| Alabama | 1 | 0 | 4 | 14 | 8 | 4 | Not Collision with Motor Vehicle in Transport (Not Necessarily in Transport for 2005-2009) | Dark - Not Lighted |
| Alabama | 1 | 0 | 2 | 29 | 10 | 13 | Not Collision with Motor Vehicle in Transport (Not Necessarily in Transport for 2005-2009) | Daylight |
| Alabama | 1 | 0 | 1 | 28 | 4 | 99 | Not Collision with Motor Vehicle in Transport (Not Necessarily in Transport for 2005-2009) | Unknown |
There could be many factors that come with car accidents, not just daylight saving time. It will be hard to completely deduct that the reason of a traffic accident is from the hour change. Our working assumption is that the only factor is the daylight saving time because we are only looking at the daylight saving time factor in regards to traffic accidents.
Another potential problem is that this dataset is from 2016. There is always an expiration date on datasets. So the question of trusting datasets from almost 6 years ago comes up. In addition, since the COVID-19 pandemic, a lot has changed, who is to say that traffic laws or how one is affected by daylight saving time have changed for Americans since.
The method we plan to use is to plot the light condition in 2016 and compare it to the plot and number of traffic accidents in 2016. We hope to find the variance and plot the variances to find a relationship between traffic accidents and light condition, to compare between daylight saving time and standard time. Our scope will become more detailed in the time frame right before and after changing the clocks from standard time to daylight saving time and vice versa. From there, we hope to use machine learning to figure out if less sunlight can affect traffic accidents and if there was a year that experiences no daylight saving time will contribute to more or less traffic accidents.