by Joanna Ngai
Communicating information through visual representation is a great way to make the data engaging and easy to understand. Here are some common mistakes to avoid to help you make the most of your data viz designs and get your point across.
A prime example of misused charts is the pie chart.
There are several arguments against using pie charts for certain data sets, whether the numbers in each category have minute differences or you’re working with a complex data set (with >5 categories). Additional categories will require more colors, legends and labeling complexity that clutter comprehension.
Our brains don’t do well reading angles of circles or comparing areas of different slices, so while pie charts are visually appealing and simple to create, use them correctly and preferably with simple fractions (ex. 25%, 75%). In most cases, finding another type of chart (ex. bar chart) to communicate your information works just as well.
This topic is in reference to 3-D effects and other visual design choices that don’t add any value to understanding any information of the given chart. These details at best, are minor distractions and at worst, end up distorting information.
Adding eye candy for its own sake is a disservice to a user trying to understand complex information.
Another example would be the radial bar chart. The problem with them is that the bars all have different radii, so bars on the outside will always take up disproportionately more area than those on the inside. Therefore, the only way to differentiate bar size is by angle, which is more difficult to do than with a chart that has a consistent axis to reference.
The exception to certain types of graphic details are when reading and remembering exact numbers is not the priority — such as when the authors are more interested in communicating the trend and the message of a piece by making it more visually appealing, engaging and memorable rather than assist with pure data analysis.
While charts are tasked to do a lot, it can be tempting to try and communicate too much info and confuse the reader. While it may be challenging to present more complex information, it is essential to prioritize information for understanding at-a-glance and remove any possibility of ambiguities.
To learn more about data visualization design:
Data Visualization Effectiveness Profile
The Visual Perception of Variation in Data Displays