Past, Present, Future: Usability Testing at the National Library of Medicine
A small number of staff have benefited from conducting our own usability testing for our NLM main website using the testing method popularized by Steve Krug and the GSA program. We have found this so useful that we are looking for ways to expand use of the DigitalGov User Research Method among our staff.
Our rationale for usability testing
Our efforts in usability testing led us to look for low cost and practical options. DigitalGov’s usability testing workshop provided an opportunity to learn new techniques to implement and share with others at NLM. At the workshop last month, the presenter discussed facilitation best practices and ways to communicate with the participant in objective ways. We learned that 70% to 80% of major usability issues can be uncovered via regular tests with small groups, even five to seven test participants.
After a brief introduction, we were able to conduct tests on two different federal staff members and act as test participants. Testers used scripts they created beforehand or scripts that were provided by the DigitalGov hosts. During each test, facilitators took notes and asked participants to perform three tasks. After our test practice, the presenter went over usability testing tools and equipment, reporting, and application of results in greater detail. The workshop equipped us with a practical model for usability testing at NLM.
Our plans for future usability testing
In the past, NLM has only infrequently made use of usability testing. As a result, website redesigns have become large, unwieldy projects. Our goal for 2015 is to apply the principles of agile development to our website refresh process. Hallway testing will be a key agile development tool because it has a low barrier to entry. Some of the advantages of hallway tests include:
- **Few resources required—**A hallway usability test can be conducted using a single staff member, a table, a single laptop or smartphone, and five to seven individuals. Steve Krug asserts that most showstoppers can be identified by a pool of five participants.
- **Rapid implementation—**A script of tasks is required to conduct a usability test. This script may only involve two or three tasks for the individual depending on the number of features of your website that need evaluation. Ideally, the test will take only 10-12 minutes, to make the testing process painless for participants.
- **Rapid results—**As only five participants are required to find most showstoppers on a site, it takes only an hour to conduct all the testing.
- **Evidence-based—**Because testing is conducted identically for each participant, the resulting data is more reliable than developers’ assumptions and guesses about which changes will improve the site.
Because the National Library of Medicine faces budgetary and staffing constraints, we look forward to using hallway testing as a low cost yet efficient way to incorporate usability studies and agile development into our website redesign process.
Developing A Community of Practice
The GSA Usability Workshop was a great experience for all our team members, so much so that we’re bringing the organizer, Jonathan Rubin, here to the National Library of Medicine to conduct another workshop for our growing community of practice. It turns out there are a lot of people working at the National Library of Medicine that are interested in assessing usability; our workshop “sold out” in about four hours, if that’s any indication. And as the NLM moves to increasingly embrace an agile and collaborative development environment, we suspect more people will begin implementing hallway testing and automated, unmoderated (online) usability testing. With new tools and a growing community of usability experts, we hope our community of practice will grow and reach products in development throughout the NIH.
Joanna Widzer, Katie Chan, Ray Bryson and Dan Wendling are librarians on the Web Team at the National Library of Medicine.
