Data Visualization through the Eyes of Color Vision Deficiency (CVD)

A Different Kind of Challenge…

My office is a chaotic (yet I maintain organized) mix of Post-it notes, old sheets of paper, notebooks, and a whiteboard covered in scrawled lists and tables - not to mention a pinboard similarly overflowing with even more paper.?These “documents” often evolve into or come together to form idealized reporting services, best practices, and ideas for deploying technologies to simplify project performance insights (and a few doodles for graphics I just need the right opportunity to trial!).

?

One topic that comes up regularly in my work, and that I’ve discussed with several reporting teams in different organizations, is how to optimize data visualizations for color-blind individuals. Accessibility is key to producing meaningful, actionable Management Information (MI)—and maybe I’m starting to feel a bit guilty for laughing off color-blind family members who’ve asked what color something is, or talked about how difficult some media can be to interpret!

?

Understanding Color Vision Deficiency (CVD)

?

Color blindness, or Color Vision Deficiency (CVD), affects around 3 million people in the UK, or 4.3% of the population. It’s inherited via the X-chromosome, affecting 8% of men and 0.5% of women. If you remember your GCSE biology, vision relies on cones and rods in the retina. Most people are trichromats, with three cone types absorbing red, blue, and green light to see a full spectrum of colors. With CVD, one or more of these cone types doesn’t function ‘normally’.

?

There are eight main variants of CVD, broadly categorized into Monochromacy (no color vision) and weaknesses or blindness in blue, green, or red perception. I’ve created a variant of the above chart below showing how each CVD variant would appear compared to normal vision:

?

A Hidden Superpower? It’s not all bad news. A study by the University of Cambridge found that people with CVD can sometimes distinguish between colors that appear identical to trichromats. This might explain reports from World War II that color-blind observers were better at detecting camouflaged objects.

?

So, How Do We Make Reports Accessible for CVD?

?

Color blindness is problematic when individuals need to compare or differentiate between multiple colors, especially when those colors carry connotative meanings (like in RAG statuses or chart labels). Below are the standard charts I will start with:

?

Here are some strategies to ensure your visualizations work for everyone, viewed through the lens of monochromacy as the most severe CVD variant to work around:

?

1. Color Palette:

Choose no more than two primary colors (e.g., blue and red) and apply varying shades. This reduces misinterpretation while maintaining distinction for color-blind users, however, it is far from the optimum solution and requires you be mindful of any visual design standards in your organization that might limit adjustments.

?

2. Shapes:

For visualizations like line graphs that don’t lend themselves well to color changes, use different shapes or icons as data markers to distinguish between data series.

?

3. Patterns:

If you’re restricted in your color choices, use patterns to fill graph segments. Adjusting overlaps or “exploding” pie charts can also help create clear boundaries. In line graphs, 'dashing', dots or other patterns can be used to differentiate between series.

?

4. Text:

Adding data labels that clearly state the series or category along with the value is a straightforward way to improve accessibility.

?

In summary, there are simple tools you can use to make your reports and data visualizations more accessible. Understanding your stakeholders and gathering requirements is crucial. As automated reporting tools like PowerBI become more common, with options for in-browser color contrasts and tooltips, the need to manually adjust visuals should decrease. But keeping CVD in mind when designing analytics, dashboards, or reports is always a good practice.

?

Hopefully this late, renewed interest in making visual communications tools work for colorblind individuals makes up for my previous lack of sympathy!