SPRINT 5: VISUALIZATION

THE PROCESS

We focused on visualization this week-- we had to find out how to present massive amounts of data into more understandable pieces for particular users in order to answer their questions. For my user (a driving tourist), the data was overwhelming and my first step was determining what information they needed to know in order to answer their question: Where can I park and when is it safe to drive?

I had to break this question into three parts: Where is there crime at night? When is there least traffic crime during the week? What time of the day has the least number of accidents? Then using these three questions, I used specific data from the spreadsheet to answer them. Using Tableau's suggested visualizations on the side provided good options that matched my data well. It was only a matter of choosing which ones were the easiest to understand.

For the locations of vehicle crime, I had originally plotted it all on a map. This was cluttered and didn't tell us where it was safe to park overnight. But instead by choosing a map that showed the density of particular crime in certain areas, it made it easier to see which place the user should avoid.

Above: The top map is messy and doesn't tell us where it is safe to park. The lower map, using various sized circles, shows us where it is more ideal to park a car overnight. It is a clearer representation that works well with our  data. 

LINK TO PROJECT: https://public.tableau.com/views/Project_141/Dashboard2?:embed=y&:display_count=yes

PROBLEMS AND CHANGES

Working with such a large volume of data, I was unsure of how to initially approach this problem. I had to present many things in a small space, and finding out which data and visualizations worked well was difficult. I had to ask myself: how do I decide which data to omit and which to include?

In the end, I had to decide which details answered the question best and would help the user in deciding where to park or when to drive. Then I had to choose a visualization that showed this the best, whether it was a graph or a map.

Next time, I would challenge myself to create even simpler visualizations for ease of understanding, or I'd include a small description in the subtitle of the data that I included. Color would also help with bringing the data to life and engaging the user.

HOW DO I MAKE AN ETHICAL VISUALIZATION? 

The ethics concerning visualizations are a tricky subject-- we need to present the data in a way that accurately portrays the data and fully informs the user. This lends the question-- whose responsibility is it to ensure that the user is not tricked? Arguably, you could say that it is the user's fault for not looking at the labels carefully and judging the data themselves, but you could also say that a skewed visualization may have the intent to deceive.

In order to create an effective objective visualization, you must first gather accurate data without a bias. The source, like with the design, can be suggestive towards one side or be wholly inaccurate. In addition, the visualization's data has to be easy to understand and objective without exaggerating any particular features purposefully.