Data visualization | Guide

First of all, don’t panic. We’re not going to get lost in technical jargon or discover the philosopher’s stone of data visualization. What I intend to do here is instead to take a general picture of the concept. Because let’s face it, it seems to be something new and complex, only for technical and analytical minds. And it is not. Data visualization has started to be so strongly related to Big Data and Artificial Intelligence that it seems to be born with these terms. I’ll let you in on a secret: it’s not. If like me, you have been confused by this situation, this article is for you. If not, it won’t hurt to take a step back and put this trendy concept into context.

Data visualization is not new. That much is obvious, and it is essentially thanks to the rise of Big Data and artificial intelligence that it has become more prominent today. But, as the primary visual beings that we are, it was inevitable that, since ancient times, we would use graphics to describe and understand the world around us. We soon discovered that our brains could decipher the messages drawn on these graphics in a more intuitive, faster, and more memorable way than with other less efficient systems

What is data visualization

Data visualization is the graphical representation of information and data. This would be the simplest definition. But, of course, there are many angles and nuances in this definition that we will try to observe here in broad strokes.

Basically, a series of tools to represent visual elements such as graphs, maps, illustrations, etc., to show patterns, trends, or outliers that help us understand the reality around us.

We immediately understood that information is power and that the more data we have, the more accurate the decisions we make based on it. So it was inevitable that more and more effective data collection tools would begin to be developed, with greater capacity to hold more significant amounts of data.

Best data visualization

Skipping a few centuries of history, we inevitably find ourselves at a point in time when, thanks to technological advances, we can collect, standardize and classify billions of data. Big Data and Artificial Intelligence automate these massive amounts of data and transform them into information and recognizable patterns. But as long as making the most compromising decisions remains the responsibility of humans, we need to understand that data and thus minimize the probability of error.

One of the keys to the issue at hand is the understanding of information. As we have already mentioned, our culture is highly visual and dependent on how our brain interprets shapes, colors, and recognizable patterns from birth.

Map data visualization

These abstract pieces ordered graphically form a code that we understand as messages if we know how to represent them. Observing the information makes it easier for us to understand and remember it, like a story.

Here’s another key. We are increasingly using data visualization to translate messages and tell stories to draw conclusions.

So think of data visualization tools and technologies as a translator. In the era of Big Data, billions of incomprehensible and unmanageable data are generated every day. Without that translator that transforms data into messages, stories built with graphical patterns understandable by the human brain, it would be impossible to highlight helpful information and visualize trends or outliers.

We tell stories about the who, what, when, where, and how using as many visual elements as possible makes the experience more memorable and understandable. Given the plasticity of the human brain and our “superpowers” of perception, accessing the data represented as a story and not a technical analysis to memorize is the best way to get the most out of the information.

But how does that insight work, and how does it help with data visualization?

Data visualization course

As designers, we know how to manipulate the power of human perception to achieve our goals. These powers draw on principles born in a very ancient and profound place in human psychology. That’s why they are so powerful, the same everywhere in the world, and so enduring over time.

Of all of them, the 7 Gestalt principles are the best known and most used by designers since they were defined back in 1920.

The term Gestalt comes from German and refers to a unified pattern, figure, form, or structure. This theory, which developed the idea that the human mind perceives things as a whole rather than as a group of elements, is, of course, still valid today.

Gestalt in data visualization

As you may have already deduced, these psychological principles are fundamental to data visualization. In fact, whether you use them consciously or not, all graphical representations used in data visualization stem from these principles and are inseparable from them, just as language is inseparable from human communication.

We identify shapes and colors that speak to us of patterns, trends, variations, differences, outliers, relationships, scales, etc. And it is in these decoded messages that our brain builds stories. An increasingly important concept in a field is scientific or technical analysis, which was previously seen as an intrusive agent coming from more creative areas.

Types of data visualization

As UX and UI designers, we should never lose sight of a design’s ultimate goal and function. We always keep in mind that less is more. Also, in data visualization. But it is not always possible to represent data simply, and we have to find the balance with the correct information.

The most common and used types would be the following. But this list will likely have to be completed with more modern formats such as augmented, virtual or mixed reality.

Charts, tables, maps, infographics, dashboards

Big data visualization systems

These are some of the most familiar specific methods for representing data. Of course, there are more, but here are many of the most commonly used ones.

And that’s all? Not at all. Most of the data we need to visualize can be represented with items from the list above. To a greater or lesser extent, we can “cook” a good story by correctly relating the information to be shown and visualizing it in the right way.

But what happens when the story gets exciting, or our client requires visualizing something new that needs a more creative and different approach? Maybe the most common graphic elements fall short, and we need to develop a solution.

No problem. Like any creative work, it requires the ability and talent to approach the problem from a different perspective, not without merit and effort.

Data visualization tools

There are numerous data visualization tools, with different possibilities for each person according to their needs and knowledge.

• Microsoft Excel. A classic Microsoft software to which we have given many uses beyond the spreadsheet. With features such as the inclusion of formulas to return sorted results, it is a tool available for several operating systems and very customizable: colors, fonts, and sorting of elements. Its plan for creating graphs and tables for data visualization is potent. It allows easy export within the universe of Microsoft Office applications (maintaining its functionality) and outside them (as an image). You can even convert tables into graphs easily.
• Power BI. This tool, also from Microsoft, takes advantage of artificial intelligence to design interactive data visualizations. It has numerous formats, from classic Excel-like charts to maps. Graphs can be customized and even created dynamically by creating formulas (DAX).

• Datawrapper. It is used to create interactive graphs that can be inserted into a website.
• Infogram. It is straightforward and visual, with many possibilities for customizing the graphics, making it an ideal data visualization tool for learners or those seeking to represent information.
• Tableau. It is another viral tool to represent data and make an analysis through the information obtained.
• Canva. This tool, with thousands of templates and images to create resources in social networks and web pages, also has a chart creator for data visualization, with many customization options (templates, fonts, colors…). Like Infogram, it is ideal for people with little knowledge of data visualization and design or who want to create simple charts. Charts can be downloaded, embedded in a website, or sent to print.

Data visualization examples

As we have already seen, Gestalt reminds us that we can play with the tools of human perception to manipulate shapes, colors, spaces, etc., until we can tell a story.

We are closer to infographics than to “pure” graphics in these cases. Think of a kitchen. We have a pantry with the most common ingredients combined in different ways, resulting in different dishes. But there are cases in which we have to look for or “invent” new ingredients, as in haute cuisine, if we want to reach a demanding public creatively and effectively.

In this sense, we are beginning to banish, at more executive levels, mainly because in other media and formats, it is very consolidated, the idea that data visualization is merely a static representation of data.

Data visualization best practices: examples

More and more interactivity and dynamism are added. We see how data changes in size, color, or position over time, how it is grouped or dispersed, how it speeds up or stops, etc.

We instinctively think of bland executive graphics in management reports. But think, for example, of simulations of star collisions, recreations of gas clouds or comet trajectories, etc. We often see this type of representation in documentaries as if it were a movie without realizing that we are looking at the visualization of billions of data, interpreted by powerful computers, presented graphically to tell a story in an understandable and memorable way.

Gaia 3D map

This is nothing new. If we think about it, we have been visualizing data in one way or another in all kinds of media for a long time. Not only documentaries but also reports, video games, TV news, … and even in publications.

Does this scene sound familiar? A news anchor tells us the evolution of the weather in a virtual scenario while showing us rainfall graphs. Another example of data visualization: we only have to wait for the next elections, for example, to see the flood of infographics about the vote count or the intention to vote that the data return.

Max Reality: Augmented Reality Weather Forecasts | The Weather Company, an IBM Business

It is becoming common to visualize data in a more dynamic and novel way. We seem to be losing our fear of talking about data from a less academic perspective, and we are telling stories.

On the one hand, this is because the audience we are addressing is increasingly prepared, and, on the other, this same audience demands a different way of accessing information.

We consume, without realizing it, an immense amount of data in video games, sports broadcasts, news programs, etc., always with an attractive visualization to the point that we are no longer aware that we are watching data but stories. Clear messages that are increasingly familiar and no longer require effort to decipher.

Another example of data visualization. Think about sports activity monitoring devices. We are not dealing with an executive report, an Excel table, or a complex management application. But we have a massive amount of data on our wrist, visually represented effectively and attractively, that helps us make decisions such as hydrate, rest, run faster, etc. And we can choose between devices depending on the amount of data we need or how creative they are when it comes to displaying it.

In fact, we go beyond the threshold of the sense of sight when, depending on whether a specific data is reached, we receive a haptic warning that helps us to make a decision: stop and rest, or you have reached your 10,000 step goal, for example.

Video games deserve a special mention because of the innovative nature of data visualization in terms of context and support. It did not take long for the need to visualize data to appear as soon as the first video games appeared, as they are inseparable from real-time information for decision-making.

3D data visualization

Seeing how data visualization has evolved throughout history and the fundamental role since its origins, it was to be expected that we would have no qualms about incorporating this tool into the most experimental and cutting-edge technologies such as augmented, virtual or mixed reality.

While it is true that the consumption of the most basic forms of data visualization would not have provided sufficient impetus to access more innovative formats, there is another world of data for which we do find the necessary motivation.

The image on the right is from Mapbox GL

We are already familiar with 2D simulations of 3D plots that help to visualize linked data on multiple axes or parameters. And we can even access many applications and websites to a 3D viewport where we can visualize the data and manipulate the perspective with much greater dynamism.

NEO Project: Virtual reality data visualization for machine learning

But virtual formats allow us to access an infinitely more immersive and interactive experience, for example, in clusters of data with different depth levels. We can move them and move between them or have a more realistic experience from the scale point of view. The possibilities are immense.

AI and data visualization

Artificial Intelligence and Big Data have a common goal: to extract the maximum value from the large amount of data generated today. But they are not the same:

Big Data refers to the storage and processing of massive amounts of data, usually collected by analyzing user behavior, to formulate predictions based on observed patterns.

On the other hand, Artificial Intelligence is a combination of algorithms formulated to create machines that mimic certain functionalities of the human brain (such as learning, reasoning, and decision-making).

How they relate to each other: Real-time graphics

Big Data acts as an input that receives a massive set of data that needs to be processed and standardizes it to make it worthwhile.

Artificial Intelligence feeds on the processed data and learns from it, creating and recognizing patterns and developing sophisticated analytics solutions.

They are not equivalent terms but often work in tandem.

Best visualization tools

And what does this have to do with data visualization? It has a lot to do with it. Because we have said that data must be converted into messages, into stories. So they must be translated and interpreted beforehand for our subsequent understanding. If we have billions of stored and processed data and an AI with access to them, we will get more accurate and efficient results to be interpreted.

This is not science fiction. There are already “homegrown” tools capable of combining access to large amounts of data with AI algorithms that can be represented graphically. PowerBI, without going any further, has been offering this possibility for some years and combining data visualization with machine learning tools and the power of languages such as Python, among others.

As you can see, we have access to potent tools that have forever changed the world of data visualization from an efficiency standpoint. But, for better or for worse, it seems that specific processes still depend on human intervention.

I have not found references to automatisms capable of selecting the most appropriate type of visualization to represent specific data. In PowerBI itself is ultimately the human who chooses one graphical representation or another. You can select a table, a bar or line chart, a scatter cloud, etc., to give examples, but the machine is not going to do that for us for the time being.

We have an ever-increasing palette of graphic options. Depending on our data and the audience we are targeting, we can choose the most appropriate way to represent the message.

But we must either build that data display system ourselves or program the machine to use a default display type. Fortunately for designers, today, a machine cannot convert this to a default display type…

In this:

Interactive data visualization

The idea that storytelling through data visualization, even though the most traditional graphics, is a success is becoming more and more solid. Not only because of its effectiveness but also because of the unique nature of the message.

But, as designers, we can’t help but remember that we carry the impulse to seek beauty in what we do in our DNA.

And it is in infographics we find the scenario that offers us the most creative freedom.

Do you think a machine is capable of performing a data visualization like this one?

Interactive infographic about Ferrá Adriá y El Bulli. Via El País Semanal

Or like these?

Author: Federica Fragapane

Author: Federica Fragapane

Infographics found their fundamental support on paper because, in the beginning, there were no computers or software capable of transferring the designer’s creativity to a digital medium. In fact, there were not even digital media.

Figurative map of the successive losses of men of the French Navy in the Russian campaign 1812–1813. Author: Charles Joseph Minard (1869)

As we already know, this has changed. Now, the only limit is precisely our creativity and imagination when it comes to translating stories and data into a balanced visual representation where form does not eclipse function and vice versa.

Poppy Field: visualizing war deaths interactively. Via Poppy Field

Although physical media retain their importance, there is no doubt that their presence in digital media is the great protagonist with all the advantages that this entails. Although it is becoming more and more difficult to access them without a subscription.

2020: The year in visual stories and graphics. Via The New York Times

Obviously, the most obvious advantages of digitizing infographics are interactivity and the ability to offer information that would remain hidden in a static medium.

This visualization brings together 123,850 letters connecting 20,424 people from the UK State Papers archive, dating from Henry VIII’s accession to the death of Elizabeth I (1509–1603). Via Tudor Networks

Another example of interactivity in data visualization.

Understanding the evolution of U.S. energy use in 1800-2019. Via University of Chicago

Conclusion

Is that all, my friends? Well, the truth is that data visualization is a book. And a big one. There is already a good amount of them, in fact, from different perspectives.

Without going into the more technical side of data management itself, the rise of Big Data or Machine Learning, the most relevant conclusion is that they are no longer simple graphical representations of information.

The revolution of data visualization, in addition to how technologies allow us to access them in a more interactive and immersive way, is the way in which this information is treated to take it beyond. To the world of narrative. But without separating it from the analytical part. Which also implies a challenge.