Bad Data Visualization Examples – Don’t make these 5 mistakes!
The drag-and-drop features of BI tools like Power BI and Tableau have made data visualization simpler and possible to use for different types of users. But, it has also given birth to many bad data visualization examples. In this blog, we take a look at all such bad visualization practices.
But, to understand why some of the pervasive errors occur, let’s first understand data visualization as a concept.
What is data visualization
According to Gartner – data visualization presents information graphically that highlights patterns and trends in data. It helps the viewers gain quick insights into data analysis.
It revolves around both data science and art. While the analytical brain is required to understand the data and analyze it, art plays a crucial role in representing it in appealing visuals.
To avoid data visualization errors, it’s required that you strike the right chord between data points and presentation. Using too many colors, incorrect charts, and presenting too much information through one graph are some of the mistakes that data analysts often make. And these, in turn, will make a decision maker’s life more difficult than easier!
The most common bad data visualization examples
1. Using incorrect axis ranges
Bar charts are very commonly used, and most viewers come to a conclusion by looking at the height of the bars. But, when the graph producer doesn’t follow the conventional method of starting the axis at 0, focusing on the height may mislead you into believing the differences as much higher than they actually are.
For example, in the given figure, the difference between the first and third bars looks much more significant than it is actually. Despite correct numbers, the impression that a truncated y-axis gives becomes a misleading data visualization.
Then, some analysts also tweak the interval between two y-coordinates, which ideally should be the same. For instance, if the next coordinate after 0 is 100, then the subsequent coordinates should be 200,300.. and so on. A y-scale with coordinates like 0, 100, 150, 250, etc., lacks regular intervals, and it might be helpful in beatification but will not solve the real purpose of a bar chart.
Correct method
- Start y-axis at 0.
- Keep the interval between coordinates the same.
2. Trying to be extra-creative
Charts and graphs used in dashboards are meant to give you a correct and precise idea of the underlying data. At times, analysts go overboard with the design and lose their focus from the standardized data analysis rules and misuse charts and graphs – resulting in the biggest data visualization mistakes.
Thus, it’s crucial to know the graphs’ purposes and when to use each one of them. For instance, when visualizing time series, using a pie chart is a bad option. Though it might look perfect in the dashboard.
Pie charts are useful when there are five or fewer parameters. Each parameter should be comparable and also have a relationship with the whole. Line charts and line graphs represent time series in the best possible method.
The given figure might look very colorful, but it isn’t very meaningful because by looking at it, the viewer will have more questions than answers. This is the perfect example of misleading data visualization.
Correct method
- Know the use of graphs before implementing.
- Avoid pie charts if the total isn’t coming up to 100 and the parameters used aren’t related.
Related content for you: How data visualization improves business decision making
3. Not using labels
Another very bad example of representing graphs is missing out on labels. Evident labels make the graph easy to understand, and as a viewer, you will have no scope of ambiguity. For instance, in Power BI, after preparing a chart, you need to turn on the data labels option from the “format” section. Since it can be easily missed out, it’s advisable to double-check on data labels.
Just implementing data labels also isn’t enough; they should be decipherable. For instance, if labels are inside the visuals, using contrast colors help. Imagine while presenting your retention rate; if the percentage is visible as 8 instead of 80, what impression it’ll have on viewers’ minds!
Correct method
- Using the graph’s title as a suitable title gives an instant gist of the data.
- Use labels that are clearly visible and easy to understand.
4. Too many colors, shapes, and texts
When correctly used, colors can set the tone and enforce a message, but too many colors clutter the dashboard. The same applies to too many shapes and lengthy texts. A map with a massive number of tiny geographic shapes, glaring colors, and lengthy text legends will confuse the reader’s mind. However, simple dashboards communicate the message more effectively.
Correct method
- Use striking colors only to highlight critical data.
- Colors have the power to evoke a host of emotions – from positivity, trust, strength, confidence, and friendliness to fear, doubt, concern, and boredom. So, while displaying sensitive information with colors, study the cultural background of your audience.
- Keep the colors in sync with the company’s brand image.
- Use images scarcely and when there’s sufficient white space on the dashboard.
- Specific colors have psychological associations. For instance, green gives the impression of positivity like profits, and red is related to losses or KPI miss outs.
- Standardize colors. If you are showing the USA in blue in one graph, keep it persistent throughout the dashboard.
5. Information congestion is another example of bad data visualization
According to a study by George Miller of Harvard, the human mind can handle seven pieces of information on average, which can be more if the information is chunked. Furthermore, a chart loaded with too much information is likely to puzzle your viewers.
And it will also be difficult for the decision-makers to filter out the information that requires immediate attention.
Correct method
- Instead of cluttering all information in one chart, it’s best to divide it into multiple charts, which together will tell the story to the viewer.
Our final thoughts…
Keeping in mind the above data visualization bad examples, your future dashboards are sure to be more error-free & impactful. And as a piece of basic core advice, always keep your audience in mind when building visualizations for them.
- Find it confusing to work around so many charts and graphs?
- Cannot really understand which visual to use when?
- Are your dashboards creating more complexities than simplicities?
Speak with one of our data visualization consultant! Fill the form below and we will get in touch.
