User Study

After we built the mid-fidelity prototype, we started running user tests with UR students from different field of studies.

We had two iterations with 5 different students each time. We used within-subject experiment (everyone uses everything) as this website is targeted for all UR students. Participants were asked to use both mentee and mentor terminals. Since the process of creating profiles for both roles are similar, the order of the tested functions would not affect the result of the experiments. We also used longitudinal study for providing a complete evaluation of the website. Participants were asked to use through the functions from registration to finding a match, which is the routine for new users.

After each user test, we had a short interview. We recorded their general experience with the prototype, asking what they think are effective and what are not, and how we can improve the existing features. All the tests were conducted via zoom meeting. Since we don't get permission from the most of participants, no video record or photo is available.

Results

We had two iterations for the user test with the wireframe prototype. The data we collected was improved after the evaluations and revisions.

After we done the first iteration, we summarized the findings and user suggestions:

• The interface is very concise, but the layout and some functions are a little confusing.

• There are some functions lacking (Searching users, leave comments, decline a mentor request)

• Users can write their intro and show their resume. They allow everyone to know each other better, and help user improve the professional identity.

• Some graduates’ study topics are complicated. We can change the selections box to a short text input.

• Every user can post their questions even if they find a mentor, and only people in related field are qualified to answer these questions.

Improvements were installed into the prototype such as color and text prompts, commenting and revised layout. Then we conduct the second iteration user test with the same method. We got new impressions and findings:

• The interface is very concise and most functions make sense.

• The matching system looks really interesting and helpful. Most of the participants are willing to use the app.

• Most liked function: Help the anonymous.