Assignment 5: Usability testing.
What a long time we spent on this assignment! Perhaps 20% of that time involved trying to pick meeting times when all four of us could show up, but usability testing does cost plenty of time to design, run, and analyze the observed data. If done right, though, the information gained will be invaluable for finding design problems that might never be encountered by the team working on the project, or even experienced heuristic testers.
As with my first assignment, we had a hard time picking a limited number of tasks to set for our users. One of these tasks, though we debated removing it, remained the creation of an account on the site, which people always breeze by. I do realize that running through that option on the site is still important, no matter how easy it seems to be to myself and to our users.
We did a fairly good job coming up with scenarios which forced the users to explore all of the relevant parts of the site. The methods we had to use to guide our users to use certain features showed me that Urbanspoon has major design flaws which really require major attention. The main search bar, for example, requires either a label, directions, or for its usefulness to be expanded beyond restaurant names and other incredibly specific phrases.
Our volunteer users testers were all college students, which gave us a limited look at the site’s usability, but nonetheless they were very valuable. We gathered diverse data sets from each performance.
Our final user seemed to be familiar with heuristics, although perhaps he was just very good at explaining his snobbery. He pointed out things that we had, for the most part, already observed, but in a discrediting tone which made me, personally, want to throttle him. This emphasized the importance of an impartial voice when reporting our testing and findings.
While we all agreed that our pilot test for our in-class usability test was very important, we opted to avoid this step in our own project because of time constraints. Unfortunately, because of this we had to do adjustments after the completion of our first test, which made… exactly the same point that our in-class setup was supposed to help us with before the fact. Ah, laziness.
But the information we gained was great for qualitative information, even though we only had time to test 4 people. It’s an important evaluation to conduct on models and current versions of products, so that improvements can be made to both prototypes and to existing things.
