Would you believe me if I told you that some people can’t tell these images apart?
It’s me. I’m one of those people. I can tell that the tritanopic image is quite different, and the protanopic image’s top light looks weird, but the normal and deuteranopic images look very similar to me.
What is color and who can see it?
According to Pilestone Glasses Manufacturing, color blindness affects approximately 8% of men and 0.5% of women worldwide. Color is something that everyone intuitively knows, yet explaining it is a difficult task. Here are the objective descriptors we can use to classify colors:
- Hue: A color’s wavelength or combination of red, green, and blue
- Value: How much white has been added to a color
- Chromaticity: How different a color is compared to a grey of the same value.
We can vary all three of these attributes to get different colors, which we can use in visualizations.
How does color relate to visualization?
Data Analytics is all about storytelling. From the data you analyze, what stories can you tell? Many times, it’s easier to explain a data trend using visuals instead of just words. Look at the following two ways to explain some data results:
“The data show that most sectors were losing jobs from 2008 to about 2010, with an average number of sectors losing jobs being around 15.”
The value of color in visuals: The example above is a visual that makes great use of color to encode relevant information.
However, here’s a simulated version of how a person with deuteranopia may see:
The contrast between the colors is significantly diminished. That means colorblind people can’t use color as well to interpret the information in graphs. In this graph, it’s harder for a colorblind person to tell how dark a shade of red or green is to determine how many sectors have a specific percent loss in jobs. So what if some of your audience members can’t see it? If you want to convince people of your ideas, they should be able to see the visuals that you want them to see. Otherwise, they may walk away from your presentation thinking: “I’m sure there were great visuals there, but they didn’t explain much for me.”
What can we do then?
The vast majority of colorblind people can still see some colors. Since most colorblind people are red-green colorblind, you can use Deuteranopia-friendly color palettes, such as those from the website Color Brewer for making maps. Furthermore, it helps to use discrete color scales instead of continuous scales, as it’s much less ambiguous to tell when one color ends and another begins. Notice the difference in the following image:
Even as a colorblind person, it’s easy for me to see 7 distinct colors using the discrete scale, while the continuous scale is harder to parse. Coloring maps this way sacrifices some precision, but vastly improves clarity.
Alternatively, you can use colorblind-friendly methods to encode information, such as orientation, shape, luminance, and density, among others.
I hope you use this blog post to ensure that your presentations and visualizations are visually accessible to all types of color vision.
Columnist: Ali Awan