To Explain, or Explore: That is the Question in Data Visualization #PoDV

This is the second post in our series ‘Principles of Data Visualization’ #PoDV

Whether you create visualizations for product dashboards, or interact with them in your day-to-day work, one way to get the most out of visualizations is to understand the goal behind any visualization. It sets the right expectations before you begin to create or analyze the visual, and gives you better results. In this post, we’ll look at two common goals in data visualization, and see how they apply to real-world examples.

Two Goals in Data Visualization

In a business context, the two main goals that visualizations have are as follows:

  • Explain
  • Explore

Let’s discuss each of these goals in detail.

1. Explain Data to Tell a Story

As mentioned in the first post, the reason we visualize data is to tell a story. If the designer of the visual has a story to tell the viewer with the data, the goal of the visual is to explain. There is a defined structure starting from the main narrative, and trickling down to each part of the visual. These visuals are effective for making a point, or conveying an insight from the data. For example, the main narrative in the simple column chart below is to highlight the country with the highest value:

column chart

The designer of this visual starts by asking a question of the data – Which is the highest value? – and designs the chart to simply answer this question. The viewer follows a similar process when interpreting the chart. This process can be understood by the following illustration:

explanatory data visualization

Explanatory visuals are editorially driven, that is, the designer leaves little to the imagination of the viewer, and crafts the visual with care to bring out the story most clearly. Presentation and design becomes important. All effort is made to reduce noise from the visual, so there are no distractions from the main narrative. Due to this, most explanatory visuals tend to be static and not interactive. This gives more creative control to the designer, and makes it easier to direct the process of interpretation.

This type of visualization is used in business scenarios for the following tasks:

  •   Answer a question. E.g., How much sales did we have last quarter?
  •   Support a decision. E.g., We need to stock more football jerseys as they were sold out on most days last week
  •   Communicate information. E.g., Revenue is on track for this quarter
  •   Increase efficiency. E.g., ‘Technical specifications’ is the most viewed section in the product page. It should be given more visibility.

Most of the visualizations we come across in day-to-day scenarios fall in this category. In business they appear in product dashboards, business presentations, training materials, and marketing content. They’re also used in the media for advertising, print and television journalism, and political campaigning.

2. Explore Large Data Sets to Discover Many Stories

Explanatory visuals are not editorially driven, but rather are viewer driven. The emphasis is not on a single important story but on discovering many small stories in the visual. The designer may not even be sure what story is there in the data. The aim of the designer is to present the data in a way that invites the viewer to notice the obvious, and discover surprising insights. It simply gives away a number of ideas for the viewer to make something meaningful out of. For example, below is a visualization of the State of the Union address of recent Presidents.

heatmap of presidents state of the union

Here, the viewer starts by becoming familiar with the visual, then identifies an area of interest. For example, ‘Which President spoke most about jobs?’ She then explores the ‘Jobs’ section of the visualization, and finds her answer – Barack Obama. She could then move on to exploring other areas of the visualization.

Exploratory visuals invite the viewer to get an overview of the visual, ask questions along the way, and find answers to those questions. The process can be illustrated as follows:

exploratory data visualization

This process can be cyclical without a specific end point. It doesn’t follow a particular order, and the viewer can find many insights, or none at all. The outcome can be to gain awareness of a topic rather than to make a specific decision. This type of visualization can accomplish the following tasks:

  •   Pose new questions
  •   Discover new areas of interest

Exploratory visuals work well when there’s a high volume of data to be visualized. The designer tolerates some level of noise in the visual to give the viewer more granularity. Because of the granularity of the data, exploratory visuals are often interactive rather than static. For example, the visual could use a drill-down feature to show or hide the various paths available to the viewer. In this case, the visual functions as an interface to the data.

Though they’re not as popular, exploratory visualizations have been gaining prominence in recent years with the rise of big data. The high volume of data, and varied data sets that have become common today can only be analyzed using exploratory visuals.

The Hybrid Model

When viewing a visual, the easiest way to tell what type of visualization it is, is to ask who does the work to reveal insights from the data. If the designer has done the work, and made the insights clear, it’s an explanatory visual. If the viewer needs to find insights that the designer hasn’t made clear, it’s an exploratory visual. That said, most visualizations fall somewhere in-between. Most visualizations are based on a curated data set that allows some, or a lot of exploration. When designing a visualization it’s important to balance both elements – explanation, and exploration.

Now that we have a good understanding of the two goals of visualizations, let’s look at a few common examples that we come across in our daily lives.

Examples from Daily Life

A Speedometer

speedometer

This visual is designed to convey just one metric – speed. There is no exploration to be done here. It’s a great example of an explanatory visual.

The London Underground Map

london underground map

The London metro map though packed with information is explanatory. It’s designed to show the best route from point A to B, and the designer has done the work to make it do just that. The viewer comes with the question ‘How do I get from point A to B’, looks at the map, and finds the route.

Blog Tag Cloud

tagcloud-example

We’ve all seen tag clouds on blogs. They highlight the most-used tags from the blog, and in that sense are explanatory, giving emphasis to the bigger words. However, they also show many less frequent tags that invite exploration. They’re a great example of a hybrid visualization.

Google Maps

google maps example

Google Maps is also a hybrid visualization. It can be used to explain the driving directions from point A to B, similar to the Londong Underground map – Explanatory. It also allows zooming, and panning to discover the surrounding areas, and landmarks – Exploratory.

Google Hot Trends Fullscreen

google-hot-trends-example

If you’ve not come across this visualization, do take a look. It’s quite fascinating. It visualizes Google search queries in real-time in the form of a colorful grid. The queries keep changing every second. This visualization doesn’t highlight any single search query, but rather invites the viewer to explore any part of the visual to find out what users are searching for right now. The story is left to the viewer to discover. This sort of visualization is possible only with the scale of big data. It’s a great example of an exploratory visualization.

In summary, the two main goals of visualizations are to explain, or to explore data. Most visuals are a hybrid of both goals. When designing a visual it pays to decide at the outset what balance the visual should have – more explanatory, or more exploratory. This single technique can greatly improve the quality of visualizations, and result in even better story telling.

This post is the second in a series of posts. If you’d like to get the low down on what’s in this series from start to end, read our white paper ‘Principles of Data Visualization.’ You don’t even have to fill in a download form to read it.

In the next post, we’ll dive into the mechanics of how we process visual information. We’ll understand the role of memory in perceiving visual information, and how to apply that understanding as we work with visualizations. Stay tuned!

Leave a Comment