June 20, 2005

5-second tests

I think I neglected to post about Christine Perfetti's article, "5-Second
Tests: Measuring Your Site's Content Pages

The latest issue of UIEtips has more from her. I hesitate to post her comments here, but since they aren't available on the UIE site, I'm going to do so and in their entirety.

Valerie Quinn of Webmethods.com asks:

Enjoyed the article on 5-second tests, but I can't figure out the
following sentences:

One of the 5-Second Test's biggest advantages is how quick it is.
When evaluating the Donation page, each user took only 10 minutes!

How did we go from 5 seconds to 10 minutes?

Christine responds:

We call the technique a "5-Second Test" because users only have 5
seconds to view each content page. However, each user spends a
total of approximately 10 minutes with us -- we include several
pre- and post-test activities as part of the testing protocol.

Before we give users a task and ask them to view the content
page, we conduct a pre-test briefing, giving them a set of
instructions for completing the test. We tell them we'll only
display the page for 5 seconds and instruct them to try to
remember everything they see in this short period.

Following the test, we also conduct a debriefing with each user.
We ask them to jot down everything they remember about the page.
Then we spend some time asking them questions about the page they
viewed, In our testing of the Red Cross Donation page, we asked
them, "What is the most important information on this page?" and
"How would you go about donating to the Red Cross?" In all, the
entire test including pre-test briefing and post-test debriefing
takes 10 minutes.

We also heard from Ted Penberthy who questions whether a 5-second test would work for a page with over 40 items of content:

On the Red Cross page, you were testing seven items. I'm working
on an application that has about 40. Too many for a person to scan
in five seconds. Any suggestions?

Christine answers:

You should think about conducting a 5-second test when a page has
a single purpose, such as a product description page or a
change-of-address page. The goal will be to identify if the design
of the page clearly communicates why it's useful.

In the case of the Red Cross Donation page, it wasn't essential
that users recall all seven donation methods. Rather, the primary
goal of the test was to uncover whether users understood there
were several different ways to donate to the organization. When
you evaluate your page with 40 items, you'll want to assess
whether the design clearly communicates the page's purpose -- not
whether users recall every item.

The 5-second test is just one technique for evaluating a design
and will not be appropriate for all types of pages. Before using
this technique with your application, you should determine the
main purpose of the page. If your 40 items serve one main
function, the 5-second test should work well. But, if the page
serves many different functions, the test results will not be very
meaningful. If the page you're working on is multi-purpose, it
will make sense to use a different usability technique, such as a
traditional usability test or an inherent value test. (You can
see more information on how to conduct an inherent value test at
http://www.uie.com/articles/inherent_value_testing/. )

Posted by bullwink at 4:55 PM