Now that we have a feel for the data, let’s go ahead and dig a little deeper into analysis at the state level.
On page 45 of the FARS manual you will see state codes. These will be useful soon.
Of course, we would hope that over the years states recognized that there were higher accidents and deaths due to drivers under
the influence of alcohol, and thus tried to do something about it. Much of what has been done over the past 45 years has been
through awareness, police force training, public messaging and stricter laws.
You will need to cozy up with the data once more and answer the following questions. All of your answers will need to be coded
in the Jupyter notebook using Python and Pandas:
(a) From 1975-2020, what was the average (mean) rate of fatal accidents which involved an intoxicated driver? This would
be over all states.
(b) In 1975 which 5 states had the highest rate of fatal accidents involving an intoxicated driver? Which 5 had the least?
Please list the states and the rates in a table in the notebook.
(c) By 1990, how much had the top and bottom 5 changed (if at all)?
(d) What was the average (mean) rate in 2020?
(e) Plot a graph with the top and bottom five states, showing just the rate over time (from 1975-2020). You can plot these in
two graphs (top 5 and bottom 5 do not have to be in the same graph).
(f) What is your interpretation of the trend – pretend you have no knowledge about the changes in law, changes in BAC
thresholds or changes in public messaging about DUIs.
