Target Users and Problem Space

We identified two groups of users: teachers and students of the National University of Singapore (NUS).  This decision was made after consideration of the prevalence of online tutorials and lectures in NUS, especially after the recent Covid-19 pandemic.

After brainstorming, we broadly identified the problem to be a lack of effective engagement between teachers and students in online classes, hosted on video conferencing platforms.

To better understand this problem, we further broke it down into four specific sub-problems:
1. Lack of built-in tools for in-class engagement with students e.g. extensive polls, kahoot-style
2.  Lack of built-in tools for students to raise questions to teacher
3.  Teacher feels like they’re just talking into a screen
4.  Screen not shared/Teacher muted situation

We analysed three other frequently used online learning/video conferencing platforms (Zoom, Archipelago, Gather.town) to gain a better understanding of the systems in place, and identify pain points or areas to target our design.

Data Gathering

After a few discussions and iterations, we refined our data gathering plan and crafted key questions for our two target user groups:

We decided that using different data gathering techniques for each user group would be most effective (based on intended sample size, characteristics of users, question type):

Data Analysis

The main question that we needed to answer about our target user group and their current practice was: ‘Is there a demand for more effective engagement tools within video-conferencing platforms during online classes? Why or why not?’

We found that the answer to our question differed depending on the type of class, and the extent of the demand for effective engagement tools varied depending on the type of users and their motivations. Effective engagement was considered highly important for seminar or tutorial-style classes, but less crucial for lectures. The most important thing that we discovered was arguably that there was a strong need for ease and clarity. From our interviews, all of our interviewed teachers have had experiences with screen-sharing issues, or forgetting to unmute as a result of uncertainty in using the interface. We also concluded from our questionnaire that while there is a significant demand for greater effective engagement during online classes, students still wish to have the freedom of choice when it comes to privacy and comfort e.g. unmuting, switching on camera.

Requirements Definition

We developed personas based on our user research and data gathered.

Two basic user personas to represent the teachers and students, respectively.

To move from data to design, we wrote scenarios for our user personas. This provided us with more vivid details off which we could base our design proposal. We crafted a current scenario to illustrate the frustrations that our users were facing, and then a proposed scenario to envision our users' future experience with our designed system. Here is an excerpt from our “current scenario":

Professor Henry joins one of the breakout room groups to check on them, but no one seems to be talking. [...] In the middle of the conversation, he is interrupted by a popup from Xiao Ming’s group requesting help and so cuts short the discussion and joins Xiao Ming’s breakout room. However, while explaining the task to Xiao Ming’s breakout room, his thoughts are again interrupted with another pop up from the other group.

And an excerpt from our “proposed scenario":

With the poll options in the toolbar, he can now set up a custom poll on the go to check on the students' progress while he waits for more students to join the online class. As students enter the call, they appear as avatars that can roam around the virtual classroom space. Meanwhile, Professor Henry positions his avatar at the front of the classroom, and there is a notification on his virtual front desk for students to open and complete the new poll.

Conceptual Design and Prototype

We storyboarded our design via lo-fidelity (lo-fi) sketches, before developing lo-fi mockups of our prototype on Miro. As the focus of this project lay in the process and UX considerations, we did not do any mid/hi-fi prototyping, instead focusing on conveying our ideas effectively via a suitable prototype and flow.

Sketched storyboard to illustrate our concepts.

The main landing screen designed for our video conferencing system.

User Testing

We focused on testing with teachers (n=4), instructing them to carry out a series of tasks that would mimic the flow of a single lesson, while thinking aloud. We followed up each session with a post-test interview to clarify our understanding of the user's test experience and ask further questions relevant to our design proposal.

Through comprehensive user testing and insights gathering, we identified the most salient issues in our interface design. This allowed us to proceed with redesigning efforts that were informed by our user's needs and behaviours.

We sieved out the issues faced by our users during testing, and developed a redesign proposal to tackle these specific frustrations.

Redesign and Heuristic Evaluation

Upon completion our redesigned prototype, we conducted a heuristic evaluation (using Jacob Nielson’s 10 Usability Heuristics). This was useful for identifying potential problems with the redesign, especially since we did not have the time and resources to conduct further usability studies. Our heuristic evaluation revealed that there were areas for further refinements (e.g. greater clarity on what the user's screen’s view is vs. where their avatar is actually located). Furthermore, if granted more time, we would rethink our conceptual model, in terms of areas that have led to greater confusion (as identified in the heuristic evaluation report).