Seeing Red

Approximately 8% percent of Caucasian males have some type of color blindness (more correctly referred to as color deficiency). The most common form is red-green color deficiency. Despite this fact, I still encounter major sites that rely on users' ability to distinguish red in order to interpret the content on a page. I've provided some examples below. These screen captures were taken a few days ago and are from major sites that receive a high level of traffic.

Example A

Example B

Example C

Two years ago, I conducted a usability study for a client whose site made extensive use of red links with underlining disabled (similar to Example A, above). Furthermore, the navigation scheme relied heavily on links embedded in text. As it happened one of the participants in this study was red-green colorblind. It was a painful and instructive session. In order to navigate, he wound up having to run his mouse over text to see which text would “display the little hand.” This needle-in-a-haystack technique (which he reported using on other sites as well) was only partially successful.

Relying on color to communicate is a pitfall that is easy to avoid, yet it still occurs. In my experience this problem, like other “low hanging fruit” usability/accessibility problems, results from well-meaning people with little or no design expertise. I emphasize well-meaning because I think the vast majority of people doing design work are doing their best (however short they may fall) to create a usable system. Still, as my father used to quote: “the road to hell is paved with good intentions.” A bit of overstatement in this context, but I do think that folks engaged in design have a baseline responsibility to understand how to make a system accessible to those with disabilities.

Update & Mea Culpa!

It's been pointed out that in my previous post I used a color-coded scheme in my sample checkout flows. In my feeble defense, the original intent of those diagrams was strictly for my personal use in collecting and analyzing data and (obviously) I didn't give much thought before posting a snipit of it online (mostly to illustrate how I was doing the analysis). The results of this research will not be presented visually in this manner--instead it will be a discussion on the amount of variability in checkout flows and where that variability occurs.

Technorati Tags: Interaction Design