For this collaborative project with Ajay Jakumar, we sought to discover how we could create an interface that was easier to use for both new and experienced customers of BART. We created a prototype, tested, iterated, and tested again.
Ajay & I created a flow in which a user can find a station, and buy a ticket to that destination. We wanted to test our assumptions, and see how we could improve the prototype.
We conducted 11 testing sessions. 6 sessions were performed without moderation, in remote user testing sessions online. The other 5 were conducted in person, with moderation. We used the remote sessions to diagnose the major problems with our original prototype, iterated on those screens, and then tested in person to get qualitative feedback.
Our research objective was to discover if the interface made the experience easier to use and understand for customers, both new and old. We noticed the points that confused testers in our early sessions, and asked them to explain the issues they were having. The most common point was understanding the map screen. Understanding which station users were at was another, and understanding when to pay were tied at 5 testers each. The next was understanding the verbiage of "two way" rather than "round trip." Generally, a greater degree of simplification was required by testers, especially in the map screen.
Ajay and I prototyped and tested through three iterations for this project. We clarified verbiage, hierarchy, navigation, and grouping for gestalt. We also simplified text and information, added affordances, and added a few features such as a search function in the map screen.
We wanted our interface to be easier for all users, both new and experienced. After our testing, we found that our improved interface was significantly easier to understand. The map screen specifically was navigated through 78% more quickly by our testers.
Ajay and I learned more than we expected about the nuances of user testing. We learned how to ask better questions, and understand the difference between what testers do and say. There is a vast difference in results from remote unmoderated testing and in-person testing, and each have their own techniques to generate better results. We are excited to use these to generate insights for our designs in the future.