6 Lessons from Human Psychology for Effective Data Visualization - Part II

In the first part of this blog, we discussed the preattentive attributes of visual perception, the power of association, why we should pay attention to perspective, and how these three elements are key when designing your visualizations.

In this second part, we will continue with the other 3 lessons that data designers should keep in mind to make the best of their vizzes:

THE WISEST CHOICE, THE BEST RESULT 

Bubble charts, funnel charts, bar graphs, bullet charts… Having so many options opens your ground to be as creative as your data allows you to be, but this can also make it more difficult to decide which one to use.

Choosing the wrong visual aid can cause confusion to the viewer as we saw with the examples above. There are some things you need to know before making your call.

Take this Stacked Bar Chart for instance:

Compare sales strategy. Source: Smashing Magazine

Stacked bar charts are used to show the composition of many items while comparing them. In this example, we can see how each of the products performed in relation to various sales strategies while comparing their sales performances. It does this by using the attribute of size to process the height of each bar. Then, it reuses the same attribute again, but this time in combination with the color attribute to see which strategies made each product perform better.

Now look at this Waterfall Chart:

Example of Waterfall Chart. Source: Best Excel Tutorial

Waterfall charts are used to show one initial value and how it is affected by other subsequent values, and the final result. In this case, a company wants to know how the overall revenue is affected by different departments, then shows the resulting net profit. Size, color, and position visual attributes are the ones we exploit in this graph.

These are just two examples of how specific charts serve different purposes and how to apply them to real-life situations. Do you want to know all the options you have and how to best utilize them? Our friends at Hubspot have an article dedicated to choosing the right chart or graph for your data.

MISREPRESENTATION LEADS TO MISINTERPRETATION

It is well known that data visualizations can be made to be vicious (GetNerdyHR even has an article on how 3D pie charts are evil). This notion is due to many news outlets and other organizations tend to use graphs that are inefficient at what they’re trying to communicate or even misleading, and sometimes this is on purpose.

Take a look at this bar graph, this time, courtesy of CNN:

Misleading bar graph. Source: Western Reserve Public Media via Statistics How To

Back in 1990, CNN used a graph similar to this one to show the political parties’ support for the court’s decision to remove Terry Schiavo from life support. At first glance, you may think that the Democratic parties’ support tripled that of the Republicans, but after you look at the numbers you’ll realize that the difference was much smaller (only 8%). This happens because bar charts need a baseline set to 0 in order to show the actual visual representation.

The next one is quite famous as it is misleading:

Misleading map. Source MediaIte

If you know me, you know how much I love a good heat map. 

The problem with this one is a bit more difficult to discern since the map is actually accurate. It is a good representation of the county-by-county results of the 2016 elections. The issue is that, as many have claimed, it doesn’t reflect the total number of voters. Although this visual is technically accurate, it is misleading at its core in what it represents. This map presents data that appears to show a correlation. They use the “show, don’t tell” rule to trick your brain into thinking there is a cause, but one trend in data doesn’t cause another. 

Just as we use the preattentive process to explain data faster and more easily to viewers, we might also use that unconscious part of the brain to cause viewers to make assumptions that may not actually be present. There is considerable debate in the data viz community about the types of obligations we have when showing and representing data. This map by itself is not to blame, but the way it is used could be misleading without proper detail and instruction. 

So many of the examples we found in studying the psychology of data visualization were of a political nature. We'll let you draw conclusions as to what this says about the power of persuasion and the responsible use of data, irrespective of political affiliation.

THE GESTALT PRINCIPLES CAN EMPOWER YOUR VISUALIZATIONS

Taking huge amounts of data and turning them into simpler, more efficient visuals that accomplish their mission flawlessly can be a hard task. Understanding the psychology behind data visualization is key to making the best of your art, and one important topic of discussion among designers is the Gestalt Principles.

Established by psychologists Max Wertheimer, Wolfgang K?hler, and Kurt Koffka, Gestalt Principles tell us that the human brain is hard-wired to identify patterns as a way to make sense of our world. Take a look at the GIF below:

They’re just moving dots. Source Gizmodo via User Testing Blog

Do you see the dalmatian? They’re actually just moving dots but your brain makes sense out of the image unconsciously. You can read more examples and a whole article from User Testing Blog focused on the Gestalt principles of visual perception, but we’ll break them down for you.

• Closure: A complex arrangement of elements are tied together by the brain as one recognizable object. This dalmatian is an example of this principle.
• Continuity: Every element in a line or curve is more likely to be perceived as related than elements that are not in a line or curve.
• Figure/Ground: We can discern the object of a scene from the background.
• Proximity: Objects that are close together are linked in our minds.
• Similarity: The shared visual properties of different objects are lumped together by our brains.
• Enclosure: The brain identifies multiple objects in a visual enclosure creating a relationship between them.

The Gestalt Principles. Source: O’Reilly

NOW GO AND MAKE YOUR ART!

Optimizing your designs to fully leverage the power of the human brain can result in some amazing pieces of art. We at Empirical love to share the best visuals we find every day to get inspired. Just for fun, take a look at this sexy viz created by Shane Mielke:

Launch It by Shane Mielke. GIF by Visme

We encourage you to go to Shane Mielke’s website to find out the story behind this techy visualization.

Use the power of the human brain, be curious, get inspired, be a creator, be an artist.

Interested in creating your own DataViz? Book a time with us if you need some help!

This blog post was originally published on the Data Lady Blog from Empirical Data when we talk about everything related to Data Science, Artificial Intelligence, Project Management, and more!