Done well, data visualization gives powerful context to the insights captured through analytics. However, there are several examples of bad data visualizations. These mistakes can lead to misunderstandings and confusion.
For example, you can see relationships between variables like GPUs, CPUs and Ethereum to make the argument that some trend (in this case, the growth of cryptocurrencies) is affecting pricing in the computing industry. We can see a relationship between Ethereum and GPUs, but not Ethereum and CPUs.
Data visualization doesn’t have to be industry-wide to make a meaningful impact on your organization. You might have data that showcases the responses to employee surveys and how those responses changed over time.
If we want data to make its intended impact, we have to be aware of the common blunders so they’re easy to recognize and avoid.
Here are a few examples of bad data visualizations.
1. Inadequate titles and labels
Even a simple bar chart can be confusing if the visualization doesn’t provide appropriate context. The graph below lacks time context and source. What data source informed this graph? Is this the average or median annual salary based on education level? Is this for a specific state, region or the United States? What year is this information based on—last year?
We have no way of knowing the answers to these questions. As a result, we require our audience to make assumptions, which could lead to misunderstandings.
2. Confused Sorting
Using the same information, we could decide to organize the categories alphabetically. However, there is an obvious way to sort education, and that would be in order from the lowest level to the highest level or maybe vice versa. When we confuse the sorting, the audience has to try harder to understand the story you’re trying to tell, which might be about the relationship between education levels and income. Confused sorting can also lead the audience to make assumptions about what the chart is trying to tell.
3. Truncated Axes
A torn graph known as (truncated axes) is when the y axis doesn’t start at zero. As a result, we can tell a story of significant change when there isn’t any. Let’s say we want to examine if there was a change in the number of page views on a blog. The actual page views for each blog post in January were 478 in 2020 and 365 in 2021.
If we show this data in a bar graph, we see that slight increase visually. However, if we change the y axis to 300 instead of 0, we see a much more significant change visually, which isn’t reflected in the actual data.
4. False 3D
When you create a pie chart that is laid back, it does two things. First, it distorts the angles, and that’s important because insights come from the angles. Second, the data on the bottom of the pie will look larger than the data at the top. So the orange and blue or the pink and green are likely similar, but intuitively you’ll interpret the orange and green as larger because of the 3D effect. Additionally, it could be easy to confuse the information in the chart, because the legend isn’t clear on which category correlates with each slice of the pie.
There are endless examples of data visualization mistakes. But, what makes data visualization good?
Successful data visualization is first and foremost accurate, but it also tells a story and shares knowledge in a powerful way. But, how do we achieve this data visualization goal?
Here is a simple checklist to keep handy:
For example, we can look at the education and income graph again. We can argue there is a relationship between these two variables. However, we’d be remiss to acknowledge all the other factors impacting both access to education and income based on career paths.
So let’s go back to and fix the graph showcasing income and educational level.
We flipped the bar graph to accommodate the longer tags. The title is clear and gives adequate context: “Median Earnings by Education for Full-Time Employed Adults in US 2019.”
We can go a step further and use notation and highlighting to direct attention and make the story we want to tell more obvious.
Do you want your data analysis to have the intended impact?
Avoiding bad data visualizations is just one way to ensure that your efforts are leveraged by leaders at your company.
Before the first chart is ever drafted, you have to make sure the data analysis is delivering answers to the most pressing questions at your company.
In Business-Driven Data Analysis, you’ll learn a reportable approach you can leverage across data projects. You’ll know how to identify the right question and the right data. You also gain the skills to communicate your findings effectively to ensure stakeholder alignment.
Author
-
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.