Story Telling With Data — Data Visualization
At the time of writing this article Google handles an average of 3.8 million searches per minute across the globe. That translates to 5.6 billion searches per day, or 2 trillion searches per year! When we look at Twitter there are 500 million tweets per day and around 200 billion tweets per year. The internet is generating huge volumes of data every day. Just by reading that short paragraph I have approached data from a numbers perspective. If I had continued with approaching more data information from a numbers perspective, I run the risk of losing your interest.
Data is however more than just numbers. There is a story, pattern and conclusion behind every data. Even more there is a way to tell a story with data and that is the focus of this tutorial. To learn how to structure your data visualization in order to effectively communicate and hit home with your data insight.
Example: Click here to view a dashboard I have made. You can hover on the images to get more insight.
The dashboard and the dataset image that has been used to create the dashboard are here below.
The dashboard is made using Power BI and the data set is a loan dataset from this website. I have however stored a copy of the same in Amazon S3(Storage folder) and gave my PowerBI permission to read that folder. I will publish an article on how to do this later.
Let’s analyze my Dashboard in a bit more detail in the following steps:
- Problem Statement — The dataset is from a housing finance company that deals in home loans. The company has a base in both the Urban, Semi Urban and Rural areas. The dataset has been generated from loan applicants and the company’s priority is to identify customers segments, and more so those eligible for loans.
- Objective — My objective for this task was to perform data exploration behind the scene and finally present my output. This involves choosing a narrative mostly focused on Gender, Loan Amounts versus other factors.
- Conclusion —
- The highest number of loan amounts are being applied by men as compared to women.
- The highest number of loan amounts are being applied by men who have no dependents, not self-employed and those who have no children.
- The lowest number of loan amounts are being applied by women who have 3 children, are self-employed and live in the rural areas.
- There is a significant drop in the amount of loan taken by both men and women when the number of dependents increase from 0 to 1.
There is so much more to conclude from the dashboard and it will all depend on your audience, what you want to achieve and the main area of focus. The conclusion statements above are what would now spark a conversation in that company as to what kind of kind of loan segments they should have and for who.
The art of story telling using data is usually guided by these key aspects:
- Understanding the context.
- Choosing an appropriate visual.
- Identifying and eliminating clutter.
- Drawing attention where you want your audience to focus.
- Thinking like a designer.
Let’s analyze the above in more detail:
- Understanding the context.
The same way a movie has a narrative, your story should have the same as well. Like a movie, your story should quickly give a general eye-opener to what the story is about, it should have a hook, and more so it should be captivating. Your story should focus on who your audience are, what they care about, and how to communicate to them.
2. Choosing an appropriate visual.
Once you’ve concluded on your context, you need to decide how to visualize your data. You will need to choose graphs that really communicate to your audience. You will also need to choose a graph that does justice to the kind of data you have. I must mention that there isn’t a single way to do visualization. Let your visuals do one thing and one thing only, and that is to meet a specific need in the best way. This is a process that will be iterative for you, a little bit of this, a little more of that. To test your visuals, ask someone to look at it and give you feedback on their observations and how they process the information.
3. Identifying and eliminating clutter.
Clutter is anything that takes up your space and doesn’t add any value to you. Treat your visual the same way you would treat a space. Too little and the space gives you an underwhelming feeling. Too much in that space and that makes you overwhelmed. In visualization, remove those graphs that are providing very little information. Eliminating clutter also involves having visual order, how you group similar observations, how you structure key points etc. You want to avoid making too many different observations or the key points will be lost.
4. Drawing attention where you want your audience to focus.
The truth is we don’t all look at things the same way. By being intentional about it, you can help your audience focus on the key things. You need to be thoughtful about how you create your visual elements to lead your audience through the information you want them to focus on. This can be achieved through things such as size, color, intensity, orientation , motion and position.
5. Thinking like a designer.
Design concept deals with three things: affordances, accessibility and aesthetics. Affordances are attributes of an object that make it obvious how the product is to be used. ex a button giving the cue that it’s meant to be pushed. Accessibility is how your visualization speaks to people of different backgrounds and skills. Your visuals could be on medical findings but your audience don’t need to be in the medical field to understand your story visuals. Aesthetics is simply how “pretty” your data is. In aesthetics be thoughtful with how you choose your color, alignment and usage of white space.
“ When you visualize then you materialize” ~ Denis Waitley
