From an early age, people like their colors. Take my 4 year old who will make it clear that pink and purple are her favorite colors.
Today, I’m going to cover a little bit about color in general but provide an emphasis on using red/green color palette.
Color is associated with many emotions.
Red for passion and power, blue for calm and trust.
We can use color to draw people in to our stories. But it’s not just the color of the data, it’s the canvas it’s placed on as well. Keying on those color associations…red for danger and blood. Black for death. The combination is striking and impactful.
We can use a shade to highlight a particular data point so that it stands out. You can see that this indie feminist comic had a good first showing compared to other comics. It’s placement is easy to discern.
We can use colors from our branding. This visualization was done for the Divvy Data challenge last year, and is done in Divvy colors. The idea is to make it consistent and recognizable.
There are several considerations for using color.. But what color does is help us convey our message …to see the data. One great association is red and green.
Oh dad, there’s so much to say.
There seem to be two major camps…Use red and green and don’t ever use it.
But I don’t think we should take this as data viz gospel. We need to think critically and ask ourselves…Who is the audience?
Are you designing the viz for yourself?
Do you know everyone in your audience? If not, your audience could have men with color deficiency. There are three types…two types of red/green (deuteranope and protanope) and one blue (tritanope). I’ll focus on the red/green deficiency because the blue is very rare.
> So what? If we use the 80/20 rule, 80% of folks would have no problems and isn’t that good enough?
Emily, last time I checked 8% is less than 20%, so aren’t we making a big deal over 8%?
Yep, because, it’s more than a standard deviation AND in some industries, anything over a 5% exception rate is too high.
Up to now, I’ve talked about the issues with using red and green and why you shouldn’t use them together.
But it’s actually much more than being considerate. If you do work for the federal government, it’s a requirement that reports, systems, and applications comply with section 508 of the Rehabilitation Act of 1973 (which was later amended).
I provided this chart because I think it was a pretty handy reference from section508.gov but one item I want to key in on is the benefits section. The higher the level of contrast colors used, the more people can see and use the content.
Associations are taught in pre-school, so they are engrained in us and very difficult to re-train. So green is good, and red is bad.
So we know there are times when we shouldn’t use red and green together.
Do not use for continuous measures.
This is the default diverging palette in Tableau. It’s effective in using color associations; green is profitable, red states have a loss. We can easily provide this extra bit of information to our visualization so that people see the message you’re conveying.
When we strip away the color association, it becomes very difficult to determine what state is profitable and which state has the biggest loss. Take California for an example. Is California the most profitable or does it have the biggest loss? We’ve lost that extra information we originally added to the viz.
So, what do you do? Use the strong association.
Chris’s visualization got me thinking, what shades of red and green can be used together? What I discovered (and is also documented in the section508.gov) is that high contrast colors work well together.
Even something bright like this can still be effective. On the surface, I would not have thought to use this bright red and green together.
But that’s where I think we use the tools at our disposal to help us make informed decisions.
The left side is the original, the right side is what it looks like to a color deficient person. Because it’s discrete and only two colors used, you can definitely distinguish the two colors on the map.
So I would say, understand what works. I was also curious to see what the color wheel would look like to a red/green color deficient person.
From there, you can pick the colors that contrast and work the best.
I’ll close with one final quote.
Be the tree.
Go out on a limb---try something new, challenging, or against the common data viz rules.
Remember your roots. Recall the foundations so that you understand WHY and can make an informed decision.
Bend, don’t break. Bend the rules, don’t break them.
I’d love to continue this conversation whether it’s now or in the future. You can tweet me, leave a comment on my blog, or engage on the facebook page for the podcast that Matt Francis and I do.
Thank you so much for allowing me to spend time with you this morning and I hope you’ve enjoyed the conference.