Resources > Articles

[Q&A] How Data Visualization Can Be Misused

Author
  • Pragmatic Institute

    Pragmatic Institute is the transformational partner for today’s businesses, providing immediate impact through actionable and practical training for product, design and data teams. Our courses are taught by industry experts with decades of hands-on experience, and include a complete ecosystem of training, resources and community. This focus on dynamic instruction and continued learning has delivered impactful education to over 200,000 alumni worldwide over the last 30 years.

woman analyzing different graphs

Alberto Cairo is a journalist, designer, and Knight Chair in Visual Journalism at the University of Miami (UM) School of Communication. He is also the director of the visualization program at UM’s Center of Computational Science and the author of several books, including How Charts Lie, The Truthful Art and The Functional Art

This article features a Q&A from the Data Chats podcast episode featuring Alberto Cairo on how organizations can leverage data visualization and how it can be misused.

 

 

Read Beyond the Title 

It’s not that charts can lie because they are just objects, but they can be misleading because we misread them. When you see a chart, don’t assume because it’s a picture, you can quickly understand it. When evaluating a chart, the first step is reading and understanding what the graph tells us. We must go beyond the chart and read between the lines. 

We can look into many different elements of a chart to assess whether the chart is adequately designed or not. For example, the visual representations being used and the level of details at which the data is represented. When we look at a graph, it’s essential to think about what the graph is trying to show us.

 

Is there a neutral choice for data visualization?

No, there is not a neutral choice the same way there are no neutral choices when we write. When we write, we need to make choices. This happens because our visualization is not just an eye representation of the underlying reality. It’s an abstraction of that reality. One of the core elements of my upcoming book is about how visualization is designed and the decisions they need to make when creating data visualization.

Therefore, any graph is an abstraction, so we need to make subjective choices. Being subjective doesn’t mean that it is not justified; we can justify the choices being made. However, every decision in data visualization can be subject to criticism and further debate.     

 

What motivated you to write How Charts Lie?

Well, I come from the world of journalism, and ideally, as journalists, we try to represent the best understanding of what the truth may be. We may be wrong, and we might need to correct ourselves. However, when I create a chart, I try to convey or represent my best understanding of the underlying patterns and insights that a data set may contain. I try to be as objective and reasonable as possible in presenting the data. 

My motivation to write the book, and my entire career, is observing how people use visualization without a purpose, such as propaganda or persuasion. And it’s not that I’m against persuasion, as it can be motivated by good and ethical reasons. But creating a visualization to sell a product, or trying to obscure the data that may undermine the perceptions to only show how great a product is, can be misleading. Those are instances of misuse that worry me.

 

Are there any egregious data visualizations that stand out to you?

One of the most egregious examples would be people at the beginning of the pandemic; in the first few months, they were using bar graphs to compare the mortality rate due to COVID-19 to the mortality rate due to car crashes. The bar graph corresponding to the speed of car crashes was much bigger than the bar corresponding to the rate due to COVID-19 because we were at the beginning of the pandemic. 

So mathematically speaking, there’s nothing wrong with that chart because you’re representing the data to compare both graphs. However, when you think about the data you are comparing, you are essentially comparing apples to oranges. You are essentially comparing something that is highly contagious and can be transmitted exponentially to something that cannot be transmitted or is contagious. So that is not a good data visualization because of the underlying misunderstanding about what the number truly represents. 

 

If you’re a stakeholder and somebody is presenting you with information, what are some of the things that you would encourage them to think about?

As data visualization is a partial representation of data, whenever we see a chart, we shouldn’t just ask about what the chart is showing but also ask about what the chart is not showing. For example, if I see a graph representing the medium, I will immediately ask about the underlying spread of the data. I would ask whether there are any outlines in the data, what the quad tiles of the distribution are and what the standard deviation needs. 

I then would ask why that wasn’t represented in the data as there may be a good reason for it. Maybe the standard deviation around a point estimate would make the chart look cluttered and busy.  

In addition, I was reviewing a graphic the other day showing what international organizations predict the world population will look like 50 years from now. A simple line chart showed the world population is predicted to reach 9 or 10 billion people in the next 20-25 years. 

My brain immediately jumped on what is the uncertainty behind that point estimate. It’s probably not a single-point estimate forecast but a range of values. This is an excellent example that shows we need to read beyond the chart of the underlying data to discover patterns and trends to communicate our point effectively. 

 

What do organizations have to learn to improve on?

There are companies I’ve worked with where I teach about good design. I teach them about good visual design and how to improve the typography, color, readability and layout. This helps to make the visualization more pleasurable and clear to read. It’s also important to instill in them the importance of making deliberate choices when designing a chart. 

The work I also do for other companies is training, so I teach people how to make visualizations and try to take their visualization skills to the next level. And not in the sense of using software tools but in the sense of improving their skills. For example, the clarity or elegance of graphics, making graphics more presentable and featuring a better design, as well as caring for the legibility of the text.  

 

Why is the acknowledgment of data visualization becoming more prominent?

I’m usually not inclined to oversell data visualization as some people may oversell it and say it can be done in minutes. And to their point, some tools help create graphics in a matter of minutes, but that does not mean the graphic will be readable. 

Also, we shouldn’t oversell the power of data visualization. Data visualization certainly can be powerful, but only if it’s used correctly. It can be used to explore data and discover certain patterns or peculiarities that may otherwise go unnoticed. So, you can explore charts in an exploratory phase of the data analytics process. And at the same time, it can be used to communicate the insights clearly. 

 

“We’ve all heard it before, ‘a picture is worth 1,000 words,’ and I try to dispel that myth because that is not true. It depends on the graphics that you are representing.”

 

What are some questions organizations can ask themselves to assess their data visualization?

Every organization is different, and not every organization needs to use data visualization at a very advanced level. Sometimes it suffices to create a simple chart in Excel. But when organizations try to identify whether they are doing good in terms of using data visualization, there is not really an abstract, general target of level or quality everyone needs to achieve. It’s best to identify this at a local level. A level that clearly and efficiently identifies with the tools and techniques you are currently using. 

If organizations believe that there is an opportunity to explore and expand their skills in communication, and not necessarily data visualization, but communication in general. Part of that effort is trying to incorporate a new tool or technique to see the kind of charts you are not using that can improve and convey your message a little bit better. And that is how organizations get better. I see data visualization as a progressive process of increased development rather than a huge investment in terms of training that the organization might not need. 

 

What are two things you would recommend organizations to do to help improve their visualization?

The first step is to sit down with the person who regularly reads your graphics and assess their level of understanding. Don’t just show them the graphic but ask what they learn from the graphic to see if you both are on the same page. If the answer to the question matches the purpose you had in mind, then you know that it’s working. But if there’s a mismatch, you know there are opportunities to learn more about their practice and try something different. 

The second step is to look at other organizations to see what they are doing. To see whether there are ideas you can borrow to implement in your daily work. Additionally, there are many resources organizations can go to get inspiration. The data visualization society is a professional association that everybody and anybody can join. The Society has a very active Slack channel where people can talk to each other about visualization techniques. 

Additionally, there are also many books out there organizations can consult. I find The Big Book of Dashboards particularly useful for business leaders because it’s not just a discussion of good practices in dashboard design. Still, it also includes many examples of dashboards from which you can get inspiration. 

 

Translate Data Insights into Business Strategy 

As a business leader, it’s important to understand what problems you can solve with data and how to leverage your findings to make better decisions. Data Science for Business Leaders shows you how to partner with data professionals to uncover business value, make informed decisions and solve problems. 

The world of data is moving fast. Understand how business leaders and data practitioners contribute at each stage of data projects to drive results that have real impact on your organization. 

Learn More

Author
  • Pragmatic Institute

    Pragmatic Institute is the transformational partner for today’s businesses, providing immediate impact through actionable and practical training for product, design and data teams. Our courses are taught by industry experts with decades of hands-on experience, and include a complete ecosystem of training, resources and community. This focus on dynamic instruction and continued learning has delivered impactful education to over 200,000 alumni worldwide over the last 30 years.

Author:

Other Resources in this Series

Most Recent

prioritizing your data projects
Article

Prioritizing Your Data Projects for Business Success: A 7-Step Process

With thousands of potential data projects to choose from, how do you decide where to start? Use this seven-step process to help with prioritizing your data projects, focusing on those that will provide the most value to your business.
Category: Data Science
Spotify is data-driven
Article

Case Study: How Spotify Prioritizes Data Projects for a Personalized Music Experience

Spotify, a titan in the realm of audio streaming, has transformed the way we experience music and podcasts. Since its inception in 2008, it’s become a ubiquitous platform, boasting a colossal user base of approximately...
Category: Data Science
Team Prioritizing Projects
Article

Avoid These Mistakes When Prioritizing Data Projects for Your Company

Even in a world full of data, business decisions often still rely on instinct and emotions. However, when it comes to business, considering all external factors before making a move is essential. This is where...
Category: Data Science
Guy celebrates connection
Article

Harnessing Data to Forge Emotional Bonds with Customers: Insights from Zack Wenthe

Zack Wenthe joined a recent episode of Data Chats to discuss the importance of understanding how consumers interact with your brand, how customers make decisions emotionally and leveraging data to create meaningful decisions.   Wenthe is...
Movers and Shakers
Article

Aligning Data Projects with Business Goals

Data-driven insights have proven invaluable for businesses, enabling them to save money and make more informed decisions. For example, UPS harnessed location data and traffic information with AI to optimize routes, significantly reducing fuel and...
Category: Data Science

OTHER ArticleS

prioritizing your data projects
Article

Prioritizing Your Data Projects for Business Success: A 7-Step Process

With thousands of potential data projects to choose from, how do you decide where to start? Use this seven-step process to help with prioritizing your data projects, focusing on those that will provide the most value to your business.
Category: Data Science
Spotify is data-driven
Article

Case Study: How Spotify Prioritizes Data Projects for a Personalized Music Experience

Spotify, a titan in the realm of audio streaming, has transformed the way we experience music and podcasts. Since its inception in 2008, it’s become a ubiquitous platform, boasting a colossal user base of approximately...
Category: Data Science

Sign up to stay up to date on the latest industry best practices.

Sign up to received invites to upcoming webinars, updates on our recent podcast episodes and the latest on industry best practices.

Training on Your Schedule

Fill out the form today and our sales team will help you schedule your private Pragmatic training today.

Subscribe

Subscribe