optimizing the office hours experience
In the lower-level computer science courses at University of Illinois, there are multiple teaching assistants (TAs) with overlapping office hours for tutoring over 500 students per course. Students use a webapp to enter a course’s single queue to meet with TAs, who physically find the student within the building and help them when it’s their turn.
finding pain points
We knew from having experienced or knowing someone who had experienced this system in the past that there was a general dissatisfaction with the system. We interviewed students and TAs from various academic and cultural backgrounds and identified pain points in the service:
poor TA-student pairings
Students couldn’t specify which TA they want to get help from. TAs and students both found it helpful to use a student’s discipline or culture as real-world anchors to relate abstract concepts to. The webapp’s pairings did not accommodate the large body of non-major and international students’ preferences for TAs who they had the best results with.
students falling through the cracks
TAs occasionally couldn’t physically find the student that is next in line due to having no way to contact the student, no way of knowing what the student looked like, and no way for the student to leave the online queue. Sometimes students would be skipped because of this.
perceptions of unfairness
Some students felt that TAs left before they understood or that other students were demanding an unfair amount of time. TA’s also felt bad if they couldn’t get to everyone who needed help.
Our interface was tested through three methods: paper prototyping, heuristic evaluation, and cognitive walkthrough. Many of our issues in our interface stemmed from lack of signifiers that communicated affordances, or signifiers that were skeuomorphic but semantically confusing.
a better learning experience for everyone
We built a high-fidelity prototype of a webapp that empowered students to get help and TAs to do their jobs.
Instead of a single queue for all the TAs, TAs now each have their own queue to avoid unhelpful pairings. Students can queue for multiple TAs at once, and are cleared from all other queues once they have started a session with a TA.
TAs can notify students through the webapp that they are looking for them, and students can reply with helpful information for the TA to find them, even if just a confirmation that they still would like help. TAs can only skip students who do not respond within a set amount of time.
To manage expectations, TAs cannot proceed to the next student unless the student confirms through the webapp that they are satisfied or until a minimum time interval passes.
Though we were able to test the interface of the webapp thoroughly, we never were able to prototype the proposed changes to the office hours service itself. Confirming that the system made sense with actors would work, but measuring the emotional and educational impact of the design would have required students and TAs with real stakes of an upcoming exam.
I did the user research, testing, interaction design, and visual design for this project.