Information Architecture Project

So this happened a while back, but it resulted in a sea change for me. I’d been working in the online/web space for ten years and growing with it, as it continued to grow.  And then, I went to Library School and spent my time being schooled on all things digital through the user-centric, organized progressive lens that is the librarian. The IA Project below, was the icing on the cake.

While I was in graduate school, the Simmons College Director of Web Design  invited the Graduate School of Library & Information Science (GSLIS) Website and Publications team to oversee the process of examining the Information Architecture of the sites. Overseen by Director of GSLIS Curriculum & Communications, the team was nearly twenty strong, and included eleven students, six of whom were doing independent studies related to the project (under the supervision of GSLIS’ Assistant Dean for Technology).

I was one of those students and this project was the highlight of my time at Simmons.

Our first task was to create user personas. The User Persona team interviewed 29 people about their use of the Simmons websites, while another 104 took an online survey of the same.

The data from those interviews and surveys were then used, along with a content inventory, to develop categories for a card sort. Ninety-three volunteers completed the card sort online.

After these first two phases, the team gathered for a retreat to sort out the draft information architecture using all of the data from the phases, as well as current site evaluations gathered by a Simmons Evaluation class. The team worked to create an architecture based on the card sort and user interviews without allowing any preconceived ideas to creep into our decision-making.

I was the co-leader of the Paper Prototyping phase where we turned the draft schema into a paper version of the website to begin user testing.  In addition to the paper prototype, we created eleven tasks for testers to complete which were based on issues found during the user interviews. Each task was tied to one of the stakeholder types that were also identified during the User Persona phase.

Before the testing began, my co-leader and I ran through the test once together to come up with an “expert” time and number of clicks to complete each task. These expert numbers were the baseline to which new data would be compared. As this was the first time either of us had done any type of usability testing, we asked our teammates to take part in a “soft open” of our testing phase. The soft open was a chance to run through the tasks using the paper prototype with a variety of users before the actual testing began. The soft open was a success and we were able to use this time to both work out the kinks in the testing process and to train the other team members.

The usability testing was split into two phases. Fifty-one users were tested during both phases, with several days in between the phases to make adjustments to any problem areas. The problem areas were easily identified; three tasks were taking longer and using more clicks than the others to complete. In between phases, we reworked the schema to make the troublesome tasks easier to finish. Since this process was about listening to the users, the “wrong” paths that most had chosen during the first phase were added to the schema as a new pathway for completing each task.

During the second phase of testing it was obvious that the adjustments had worked; all three tasks were being completed in less time and with fewer number of clicks. The success rate of each task had gone up during the second phase of testing. The User Persona and Card Sorting phases had given us good enough data so that we built a successful schema and the first phase of testing had allowed us to work out any remaining kinks. The usability testing had been a success.

My final task was learning Omnigraffle to create wireframes based on our usability testing –  to be presented to the Director of Web Design.

What the Simmons GSLIS IA team lacked in experience, it heartily made up for with dedication and hard work. My experience on the IA team debunked all of the excuses that I had become accustomed to hearing for not doing usability testing. In roughly 2 ½ months, an inexperienced team was able to engage over 400 members of the Simmons community to collect data on and test the usability of the Simmons websites. While it was an arduous process, it was not impossible to do within a short time span, with little money and no real experience to draw from.

The overall project was a success as members of the team presented our findings and proposal to the Simmons IT team and they were accepted.

I wrote my final paper for this independent study on the importance of usability testing, drawing from the words of Steve Krug, Jakob Nielsen, Christine Perfetti, Jared Spool and Lane Becker from Adaptive Path.