Data visualization is, no doubt, the practice of translating information and data into visual contexts, intending to make the data we want to communicate to our audience more accessible and easy to understand.
But when dealing with visualizations, as it happens in many other fields, there are dos and, of course, some don’ts to consider accurately to not compromise the final result. To communicate more effectively and avoid misleading visualizations, let’s give a look at these 5 common data visualization mistakes.
1. Choosing the wrong audience to present your data visualization
All too often, taking some time to study your audience is an underestimated process. But knowing your audience is crucial for creating impactful data visualization and communicating your data and information in the right way.
So, before spending hours a time making charts you won’t look forward to sharing with your audience, take a step back – even two if you need – and answer these questions: who do I want to share my data with? what is the message I want to deliver and how do I want to communicate it?
In a nutshell, even when dealing with charts and graphs, data, and numbers, you always have to focus on the audience you want to share your visualization with by taking in mind to consider the reference context.
And this is because some types of charts, for example, are used in specific fields to communicate certain information related to a given context. Among all the data visualization, we can name the candlestick chart (also called Japanese candlestick chart), a type of chart used in finance to describe the trend of the stocks.
Finally, a helpful recommendation is to show a colleague the data visualization we would like to share with our audience and ask where they focus their attention and what questions come to their minds. This is, no doubt, an excellent method to fix eventual data visualization mistakes in advance.
2. Using a bad type of chart or graph
The process of choosing a data visualization sometimes can be misleading, and it can lead to wrong choices that will make our story with data challenging to understand. The truth is, not all types of charts and graphs are good at representing our data visually.
Read our previuous article on how to choose the right type of chart
Now, the question arises, when to opt for a bar chart, a line chart, or an area chart? Even before focusing on the visualization in itself, think about the size of your data. As you may have heard, some types of charts don’t work well with large datasets because they would require an extra effort to our brain to interpret the information to communicate. On the other hand, the same chart would suit well with a smaller dataset.
So, once figured out the size of the dataset, one other aspect to consider is the type of data collected. Data can be qualitative, quantitative, continuous, or descriptive. Knowing the nature of your data will be helpful to discard some types of charts in favor of others. Cole Nussbaumer Knaflic, the author of the book Storytelling with Data, gives some practical advice. Let’s see them in detail:
Pie charts are good at showing proportion, usually in percentage, among data. Sometimes they are misused: for example, to evaluate the dimension in absolute values, comparing data among other pie charts, or for showing percentage values that don’t reach, in the total, the 100%.
Bar charts are one of the most common data visualizations. They help compare data categories, like sales for products, but they should be absolutely avoided to represent a temporal trend. In that case, the best option would be a line chart.
This is probably one of the primary rules concerning data visualization: don’t use 3D charts! The reason is pretty simple, and it is due to the fact that the third dimension chart twists numbers, making them almost impossible to compare. Moreover, 3D introduces elements that don’t add value to the data visualization but are just distractions. In short, don’t use a 3D chart if you don’t want to receive negative feedback.
Charts with secondary y-axes
It’s not always a good idea to add a vertical second y-axis on the right side of the chart because it could be less intuitive to understand what data should be read concerning an axis or another.
3. Overcrowd the data visualization of useless information
Once you have identified the audience to deliver your information and have chosen the right type of data visualization that best suits your intent, you just have to add data within the chart.
Before moving to the next step, remember that every additional information you insert within your data visualization implicates an extra cognitive effort to your audience. To summarize, every data to be understood must first be processed, and this requires brain energy.
Therefore, the more information you will add to your chart, the more you will put your audience under stress, and the higher the risk that all that data will never be understood and you will have lost your audience’s attention forever.
What’s the solution? Remove unnecessary or irrelevant details to ensure your audience pays attention just to the most critical data. In jargon, we speak of decluttering, which is the process of removing clutter from charts.
Let’s see some of the elements that make the data visualization chaotic:
bold title: it’s ok to add a headline in a visualization, but there’s no need to emphasize it with an extra bold font. Instead, opt for a different font weight so that it won’t capture all your audience’s attention.
borders and grids: consider removing them if they don’t add informational value to your chart because they can lead to a misleading data visualization.
legend: it would be better to label lines rather than add a legend on top of the visualization, forcing your eye’s audience to go back and forth between legend and data.
x and y axes: one best practice is to push them in the background and color them in grey to not compete with data.
excessive use of color: we will talk about this issue in the next paragraph, but remember that using too much color within charts is not as cool as it may look.
4. Don’t go overboard with colors your data visualization
When it comes to data visualization, color can be your best ally or, on the contrary, your worst enemy. Colorful visualizations may look eye-catching, but if you use excessive and meaningless colors in your charts, they will stuff up all your good intentions into creating an impactful visualization.
If you want to learn more we have written an article in our user documentation on how to manage colors in BStreams
5. Creating misleading visualization
One of the worst data visualization mistakes is creating misleading charts. Choosing the wrong type of chart or graph can lead to some misinterpretation with a high level of risk when it comes to making critical data-driven decisions.
The following example best explains what we are talking about. The x-axis shows a timeframe with regular intervals of one year from 2010 to 2012 and then represents the year 2013 divided into four quarters. If you don’t pay attention, you could interpret a slowdown in non-performing loans, while if the year 2013 had been shown as a single time interval, the exact opposite would have emerged.
Conclusion: avoid these 5 data visualization mistakes
To sum up, next time you will work on your data visualization, make sure to avoid the following pitfalls:
- choosing the wrong audience for your data visualization
- using a bad type of chart or graph
- overcrowd the visualization of useless information
- make excessive use of color within the chart
- creating misleading visualization
These are just some of the most common data visualization mistakes. Avoiding them will allow you to effectively communicate your data storytelling and connect with your audience.