Photo by RODNAE Productions

When newcomers are introduced to the field of data science, explanations often center on the capability to add immense value for organizations across all sectors by leveraging insights from data to drive more-informed decision making, forecasting and more.

That’s why the list of soft skills in a typical data scientist or data analyst job description often includes a line like: Ability to communicate complex ideas to non-technical stakeholders.

However, the widespread need to communicate data insights with those who will decide how to act upon them also connects to a crucial hard skill. Of course, we are talking about data visualization.

What is Data Visualization? 

Data visualization is simply a graphic representation of the derived insights extracted from the data.  Effective data visualization communicates these data-driven insights in a way that is accessible to their respective audience.  It is widely regarded as one of the most important skills in data science and data analytics.

Big picture: Data visualization “helps to provide stakeholders and other team members with quality information by transforming massive amounts of intangible data into easily understandable pictures and graphics.”

With its capacity to generate valuable business insights and solve real-world problems, data science is now driving innovation and change across nearly all industries and sectors — health care, eCommerce, law enforcement, marketing/advertising, transportation, sports, entertainment, nonprofit causes, and more.

Data visualization is the key capability that helps everyone involved as visualized data is processed faster in the brain than written text, meaning it is easier to understand, and visualize the big picture from the data-driven information.

The Importance of Storytelling with Data

No fancy charts or graphics are needed to communicate the importance of data visualization. But while data visualization at its essence involves more easily conveying key information to relevant audiences, it also relies heavily on the concept of storytelling with data which connects your narrative to your visualizations.   

When you are utilizing data, or more formally the technique of data visualization, to produce work-related reports or presentations, it helps to be mindful of the idea that you are essentially using that data to tell a story and will serve as the jumping off point for decision-making.

“Data storytelling is the general term we use to describe the full act of gathering data, extracting insights, and translating those insights into a story,” according to visual communication company Column Five Media. “It’s a compelling narrative crafted around and anchored by compelling data, used to guide decision-making, reveal an interesting trend, or provide valuable information to your audience.

At its essence: “Data storytelling is the blending of two worlds: hard data and human communication.” As such, data visualization and storytelling are at the heart of what makes data science so interesting to many of those who practice it.

Types of Data Visualization 

When it comes to practicing the art and science of data visualization, there is a wide variety of specific techniques, strategies and formats.

Tableau, the software company behind one of the most widely used data visualization tools offers the following list of the most common general types of data visualization:

• Charts
• Tables
• Graphs
• Maps
• Infographics
• Dashboards

Diving deeper, Tableau cites these more specific examples of methods to visualize data:

• Area charts
• Bar charts
• Box-and-whisker plots
• Bubble clouds
• Bullet graphs
• Cartograms
• Circle view
• Dot distribution maps
• Gantt charts
• Heat maps
• Highlight tables
• Histograms
• Matrix formats
• Networks
• Polar areas
• Radial trees
• Scatter plots (2D or 3D)
• Streamgraphs
• Text tables
• Timelines
• Treemaps
• Wedge stack graphs
• Word clouds

In addition to sharing some excellent examples of data visualization in action, Tableau also offers a look at “10 of the Best Data Visualization Examples from History & Today” — from Napoleon’s March of 1812 and London’s 1854 cholera outbreak to U.S. population trends by age group (1950–2060)

How to Infuse Data Visualization into Your Work & Your World

In terms of tips for giving data-heavy presentations — as well as practical advice for anyone looking to create and share data in visual formats — Column Five Media suggests paying close attention to three very important fundamentals:

1. Find the story in the data. What is the unique story you are trying to tell? A new opportunity? A new efficiency? A cautionary flag about a potential risk?
2. Craft a narrative around that data. It is important to guide your audience through the story by providing context for the data, creating a logical flow and making it easy to understand with interesting titles and subtitles. It is important to know your audience so that you can create a connection within the narrative that will appeal to your audience in some way.  An audience who is emotionally invested is an audience who will answer your “call to action”.
3. Choose the most effective data visualization. This is about selecting the format and visual design to maximize comprehension of your data-driven story. Once again, it is critical to understand your audience before you craft your visualizations.  You need to ensure they are accessible and easily interpretable to your audience. Visualizations that require your audience ample brain power to decipher will quickly become either ignored or forgotten.

In another post focused on how to effectively use data visualization formats and techniques detailed in the previous section, and the principles of storytelling with data cited above, Column Five Media also offers helpful advice on creative ways to present your data visualizations. The list includes eye-catching examples of each:

• Infographics
• eBooks and white papers
• Enhanced traditional reports and presentations
• Interactive graphics
• GIFs and animated visuals
• Motion graphics and videos

In terms of selecting the right format for the right data visualization application, business intelligence software firm Sisense goes in-depth on 13 different data visualization options, from basic to more advanced, explaining the unique features of each and when to use them for maximum impact. Their list also includes visual examples of the following formats:

1. Indicators show one KPI, clearly
2. Line charts display trends
3. Bar charts break things down, simply
4. Column charts compare values side-by-side
5. Pie charts clearly show proportions
6. Area charts compare proportions
7. Pivot table easily present key figures
8. Scatter charts: distribution and relationships
9. Bubble charts: understand multiple variables
10. Treemaps display hierarchies, compare values
11. Polar charts show relationships between multiple variables
12. Area/scatter maps show geographic data
13. Funnel charts display a pipeline, typically for sales figures

Finally, in an article oriented toward those less familiar with the topic, Harvard Business Review gets to the heart of why it’s so important to have a strong grasp on the key principles and practices of data visualization. Simply stated: “How you present data can double — or decimate — its impact.”

 
 
Erin Cooke is currently an instructor in “Data Science for Good,” a special workshop series produced by the University of San Diego Master of Science in Applied Data Science program in collaboration with San Diego Tech Hub. She feels a special privilege with this opportunity of introducing the field of data science in this program and touching upon the important subjects of Data Ethics and Storytelling with Data.