Heeding the users’ voice

October 15, 2014

First, people invented the Internet. Then, people started getting lost, chasing the needle in the Internet haystack.

Anyone who has ever been lost on a website quickly appreciates the value and art of usability testing. It’s a crucial step to creating a website that is intuitive, user-friendly, easy to navigate, and simple to learn and remember.

At Statistics Canada, usability testing of our website is an ongoing endeavour. It is an important one because StatCan has a daunting amount of material—the largest information trove among all government departments and agencies.

Gazillion megabytes

When you have a gazillion megabytes of data, you do not want users to feel like they are on a treasure hunt. Helping users quickly find the right information nugget is important.

Are we perfect? Not yet...but we are working on it.

Nancy Hofmann, head of Client Consultations, leads a small team of usability specialists who work to ensure that website development at Statistics Canada is not just about the technology and apps, but also about the people using them.

“We see ourselves as the users’ voice,” Ms. Hofmann says. “Statistics Canada has a wide range of data users. Developing a website that meets the needs of all those users is challenging.”

Next month, the client consultation team will play a central role when Statistics Canada launches a beta site to test a new way to publishing our information and asks users to test drive it. The team will compile user feedback—good and bad—and provide independent testing results to the web designers and developers on what works and what needs fixing.

Keep watch for the beta launch in mid-November. The more users provide feedback, the better the site will be. The better the site is, the easier it will be to use. Then, those users can provide feedback again. This way, the users can drive usability.

Testing room

Back to usability testing and how it unfolds at Statistics Canada. The consultation team has its own lab—a small conference room with a large screen and recording equipment. The room is connected to another room where team members can observe individual test sessions.

Generally, in individual sessions, participants are given a series of tasks, such as finding the population of a small town in Saskatchewan. Then, they are asked to click their way to the answer.

Why the recording? “What is important is not just where a participant clicks, but we also watch to see their reactions,” Ms. Hofmann says. “We encourage them to be as open and honest as possible and to talk out loud so we can understand their thought processes. What people sometimes do not realize is that it is not them who are being tested. It is a test of the website.”

A note-taker records where participants click on the website, and takes note of their frustration and/or satisfaction levels.

Card sorting

For another test, known as card sorting, users are given a set of cards with labels and asked to categorize them. The labels that make sense to the average Statistics Canada veteran may not make sense to the world beyond. “Our StatCan jargon is not always well understood outside our building. We understand our data, but we have to make sure outside users understand how to find our data,” Ms. Hofmann says.

The team tries to draw in a wide range of users—frequent and infrequent, experts and novices, statisticians and students, mobile and desktop users.

It is also true that what works in English may be problematic in French. Usually, a usability lab test of eight people, in each language, helps to flag any potential problems with the site.

Focus groups

Sometimes, the team also hosts and moderates focus groups in cities across Canada. In these sessions, six to eight participants are asked to share their thoughts, and a moderator is able to probe for specific issues and to follow up on comments.

As well as the face-to-face consultations, online consultations are another way to get a reading of larger groups. One online method, like the upcoming beta-site testing, asks users to try the site and share their thoughts or rate pages.

As well, the Web Evaluation Survey is conducted annually to gauge user-satisfaction levels. In the 2013 survey, 66% of users said that they could find the information that they were seeking. As well, 69% were satisfied with the website.

Where do we find recruits? Some participants in the Web Evaluation Survey volunteer for more in-depth consultations. Sometimes, StatCan directly recruits experts who regularly download data and know a specific aspect of the website extremely well. Sometimes, people visit the Consulting Canadians module where results are posted, and volunteer to be part of a consultation.

Then, there are people who stumble onto the Statistics Canada website looking for passport information, or for the most popular pet names (we do not track these), and decide to stick around to offer advice on how to improve the website. For the consulting team, a person looking for puppy names is an excellent new prospect for our usability testing because they bring a fresh perspective to the test. In the end, Statistics Canada wants to make sure that all users can find what they are looking for.

Please note that comments are moderated. It may take some time for your comments to appear online. For more information, consult our rules of engagement.

Date modified: