In data visualization, a good report is one that readers can easily interpret. Essentially, if you design a report that is unsightly, confusing, or hard to read, you’ve failed at your task. One of the primary devices we use to make our charts and graphs more readable is color – it should come as no surprise that reports devoid of color are visually uninteresting.
Almost all reports use color, but not all reports factor color-weak or color blind into the design. This is usually an oversight. If you don’t know any color blind people, it can be easy to forget that some readers might struggle to read your charts.
Although the figures vary somewhat depending on where you look, it’s generally agreed that around 1 in 12 men (8%) and 1 in 200 women (0.5%) are color blind(stats from Iristech and colorblindawareness). That equates to over 350 million color blind people worldwide – not far off the population of the US!
As you can see, color blindness is far more common in men than women. This is because the gene for the most common type of color blindness is carried on the X chromosome, and since women have two Xs, a defect in one can be compensated with the other. By contrast, men only have one X chromosome, so they don’t have this luxury, and the gene will always be expressed.
First, it’s important to understand that color blindness doesn’t mean someone can only see in black and white. That is called monochromacy (where the person can only see shades of black, grey, and white) and is exceptionally rare. The most common form of color blindness is red-green color blindness (deuteranopia), affecting most of the color blind population. People with red-green color blindness have difficulty distinguishing different shades of red, green, and yellow.
We can also break this down a little further. Some people with red-green color blindness will be primarily deficient in red color vision, while others will be green.
Someone with primarily green deficiency (deuteranomaly) might confuse mid-reds with mid-greens or mid-brown, bright greens with yellows, blue-greens with mid-pinks, pale pinks with light grey, or light blues with light purples.
Someone with a primarily red deficiency (protanomaly) may confuse blacks and reds, blues with reds and purples, mid-greens with oranges, and dark browns with dark reds, greens, and oranges.
People can also be blue-yellow color blind (tritanopia), although this type is far less common, affecting approximately one in 10,000 people.
If you want your charts and other visuals to be understood by everyone, then you need to design with color blindness in mind. The good news is, doing this is less complicated than you might think. Just follow these simple rules:
For people with red-green or blue-yellow color blindness, the more colors you use, the harder it will be to tell them apart. This is also true for non-color weak readers – the more colors you use, the more you rely on similar shades of the same color.
One way you can get around this problem is by opting for a more color-minimal design. This doesn’t mean you need to remove color altogether, but rather be selective about how many colors you use. For example, you could use color blind-friendly colors to highlight the most critical trends on your Power BI graphs and charts.
In other words, don’t worry too much about creating six, seven, eight color blind safe colors because you probably don’t need that many colors anyway. Sometimes less is more.
But let’s say you need to use many different colors in your visuals; what should you do then? Use a color blind safe palette, like this one:
This palette offers eight different colors that people with both protanomaly (P) and deuteranomaly (D) will be able to distinguish.
With that said, we always recommend opting for a more minimal design across your reports, so you should ask yourself whether you’re including more data than is necessary to communicate your point. Beyond color, too many visuals detract from the message. Not only are minimalistic designs timeless, visually pleasing, and help draw attention to key insights, they’re also better for color accessibility.
If you don’t want to be restricted to a color blind safe palette, you can improve accessibility by using colors of varying lightness. You might have heard the phrase “get it right in black and white,” and that’s what this tip is getting at. Essentially, if your charts are still readable in black and white, any color blind person will also be able to read them. Following this rule ensures that you create contrast in color lightness rather than hue – hues are difficult for color blind people to distinguish.
Rather than using legends, label your graphs instead. Labels add clarity and ensure your readers don’t have to read your graphs in two steps, and they also ensure color blind people aren’t misled.
Additionally, if you expect your readers to use color names when discussing your charts, you should clearly state the name of the color. Remember, recognizing two colors are different isn’t the same as identifying which colors they are.
Relatedly, you also shouldn’t use color alone to signal information. For example, non-color blind people typically associate red with “warning” or even “bad,” but a color blind person is unlikely to come away with the same understanding from the color alone. In situations like this, you should consider using symbols for clarity.
Using thicker lines in your charts can make it easier to distinguish between different colors.
If you’re unsure how your Power BI reports stack up, it’s a good idea to run some tests with color blind simulators like this one. You can also use tools like Colorable to test differences in color hue and lightness and find color-safe combinations. Crucially, Colorable will even give you a pass or fail score based on the W3C (The World Wide Web Consortium) Recommendations.