|paper prototypes for livingvotersguide.org. photo by Jonathan Morgan|
Making the web more user-friendly, one test at a time
I have a rich background of professional experience in user research and usability testing. I am equally proficient with quantitative and qualitative approaches: I have performed task-based usability studies, semi-structured interviews, and statistical analyses of log files and user metrics. I perform user research and usability work at all stages in the design process using personas, paper prototypes, alpha-stage interactive prototypes, and fully functional web applications. I specialize in the analsysis of online communication and collaboration tools, where subtle design features can have a profound impact on the usability, desirability and ultimately the success of the product.
I understand that the results of a usability test have to be usable, too. That's why I excel at creating pointed presentation of test results that are easily understandable and resonate with other members of the project team. My own design experience allows me to make detailed and directed design recommendations based on concrete observations and professional expertise.
The Living Voters Guide
On November 1st and 2nd 2010 we conducted seven lab studies with paid participants recruited from the gigs section of Seattle Craigslist. This goal was to (1) get a sense of the perceived value and relevance of the LVG before the election had passed; (2) learn how well users understood the basic interaction mechanisms of the LVG; and (3) understand how participants reacted to the points other users submitted.
To achieve these goals, participants were asked to interact with the Living Voters Guide for a period of 35-40 minutes while thinking aloud. During the study, a researcher sat beside the participants, observed their activities and asked questions occasionally about their actions and motivations. The study was minimally structured: participants were not briefed beforehand about the nature of the site, and were free to interact with the site in whatever way they wished. Only occasionally were they prompted to undertake certain actions when the researchers desired feedback on specific functionality that the participant was consistently bypassing or appeared unaware of. Participants' spoken observations and on-screen behaviors were recorded for subsequent analysis.
I was asked by the makers of DiscoverText to evaluate their commerical tool for collaborative analysis of online text-based content (such as Twitter tweets and Public Comment forum messages). The designers were interested in an end-to-end evaluatino of DiscoverText, from creating accounts to gathering, annotating and analyzing data. They were also interested in gaining a better understanding of the collaborative aspects of the tool. For this evaluation, I recruited users in three sets of two, and asked them to perform a comprehensive series of basic individual tasks, as well as performing collaborative tasks in which they interacted through the website while sitting in separate rooms.
Reflect is a new tool designed by researchers at the University of Washington that makes a simple change to the familiar comment board interaction paradigm: allowing commenters to insert concise inline summaries of others comments. The tool is intended to facilitate reflective listening behavior on the part of the summarizer by making it easy for them to synthesize points made by others. Reflect is also intended to enable the highlighting key points by allowing readers to draw them out of the comment thread and list them on the sidebar. Reflect summaries are designed to make it easier for other readers to scan through long threads and get the gist of what has already been said, without having to read every word.
For this study, I used a think-aloud protocol with four participants. Participants were assigned to perform a variety of tasks on a threaded conversation that had Reflect summaries enabled, in which some (but not all) comments had been summarized by other users. The designers wanted to know if people would 'get' what Reflect was for, whether they would understand the purpose and value of the tool, and how they would choose to use it without external prompting.
Virtual Concierge Widget
I was asked by Bedyamic Inc. to evaluate a new product called the Virtual Concierge Widget, a web-based Flash application for helping travelers find interesting things to do while visiting an unfamiliar city. The widget allows travelers to browse events by category (such as Theatre, Music, and Kids and Family), add the events they are interested in to an itinerary, and then print out or email that itinerary for future reference. Bedynamic particularly interested in the following aspects of the user experience:
- ease of finding events of interest
- ease of selecting events, purchasing tickets and creating an itinerary, and
- ease of printing or emailing itinerary
I performed this usability study from start to finish, six volunteers who had no prior experience with the product. I created a list of realistic user tasks, facilitated the usability study and wrote up a report of the salient results, with an emphasis on the primary user pain points and trouble spots. I also took quantitative measurements of how successful users were in completing the assigned tasks, and also devised a post-test survey for users to rate their experience on a variety of qualitative metrics and provide useful feedback.