Introduction

I conducted a small experiment to determine whether hue and value together can be used to differentiate visited and unvisited link colors on a web page, or if value alone is enough.

Method

• one condition used green as the visited link color.
• one condition used light red as the visited link color.

The participants were all experienced web users. There were a total of 20 participants - 10 in each condition.

I emailed each participants one full-sized image of a hypothetical web page. In the email, I asked them to look at the image and report which color strikes them as the visited link color. Here are snippets of the page participants received:

Condition 1: Red & Green	Condition 2: Red & Light Red

Note: The value for the green color is higher than the red color and thus can be distinguished by red-green colorblind persons.

Results

The following table shows the number of people who voted for a particular color as the visited link color.

Conclusion

Most users expect value to indicate the visited link color; they expected that the visited link color would be lighter than unvisited link color. Here are some of their comments:
- "I expect my visited link color to be more muted."
- "I figured that because it's in a more of a pastel color."
- "paler = visited"

The results show that when value alone is used (Condition 2) as the indicator, participants were overwhelmingly (80%) able to correctly identify the visited link color. However, when value and hue were involved (Condition 1), participants were not able to easily distinguish the higher value color. As a consequence, the responses in this condition were almost evenly split between the two colors as being the visited link color, and not being sure.

Caveats

There are several caveats about this experiment that I want to identify:
- Participants were interacting with an image, and not a real site. It is possible that on a working site, the users may not encounter an issue with which links are visited.
- I only tested the red and green combination. It is possible with other color combinations, users may have been more successful in identifying the visited link color.
- Because the experiment was conducted over email, I could not account for the type of monitor that participants were using, which may have affected the results.