Day 4 done, one more to go to wrap up Dashboard week. In this blog, I will go through the data, approach and end result of using car accident data to build a viz in Tableau.

The data


Today has been one of the “easier” days this week. The data we were given was clean and plentiful. The topic was accidents in Victoria and the data consisted of 12 data files, each one containing Accident ID which made it so easy to join. This was very good, given the challenge that we were not allowed to use Alteryx to modify the data.


Tableau is great when it comes to joining data, even if you have different granularities of datasets. This was the case for my data so I decided to combine my data sets using relationships. Additionally, I created my own excel file with coordinates for car impact points (so I can show where on the car impacts happen using a picture of a car) and created a relationship between this new file and vehicles.



The approach

The first thing I did was identity a specific question I had about accidents and then to go through the data sources and find all the columns I need to answer my question. This significantly reduced the number of fields I had which made it easier to focus on my question. I decided to go with car safety by manufacturer given that an accident happens. I used the level of damage to a car and the extent of injury to the people in the car (including mortality rate) and looked at which car makes are the safest. I also included a speed zone filter in the view to show the difference in damage and injuries per speed zone.


The viz

The outcome of the viz was an exploratory view where the user can choose a car manufacturer to investigate by clicking on the car make icon. The entire view then updates to show the statistics of accidents for that particular make. The user can then go a step further to choose an impact point by clicking on it, this will further update the view to show the level of vehicle damage and extent of the injury when a car is hit at that specific point.



Additional views


On top of the displayed above, I also wanted to show which cars are most popular in Victoria based on the accident data. I also wanted to show how the number of accidents changed over the years starting from 2000 – 2020. I included both these charts as viz in tooltips so. The first graph was included as a viz in tooltip over a button (star), the second was included over a viz in tooltip over the bars. This allowed the viz in tooltip to be filtered by the bar. Additionally, I brought the accidents over time graph into the view but set the coordinates of the container so it is not on the dashboard, this allows the view to be filtered by the car make filter as well.



Overall, I am pretty happy with the result. There are loads of extra features I would like to add to this view after this week is over. I am also happy to report that I time-boxed this exercise and got to bet at a normal time for the first time in three days.


Keep an eye out for my blog on Day 5 of Dashboard week, until then; happy vizzing!



Charisma Adlem
Author: Charisma Adlem

Charisma has an interesting background in animal science, having completed a Master’s degree (MSc) in Zoology at the University of Pretoria, South Africa. She found her passion for data analytics through her scientific studies. She was delighted to discover that The Data School provides a means to follow her heart and enter a career in data analytics. Charisma is a loving mother of two ferrets and has discovered a talent for abstract and realism painting in her spare time. If Charisma had to choose only one food type to eat for the rest of her life, it would be sushi. Charisma also enjoys outdoor activities including fishing, camping and hiking.