MealChat
Promoting mealtime social interaction for college students eating on campus.
This work was chosen for presentation at CHI 2020 as Late-Breaking Work.
Read the submission for more information.
Overview:
This project was completed in collaboration with the IU ProHealth Lab. We were challenged to work in cross-disciplinary teams to research, design and implement a concept to improve health.
Duration: 12 Weeks
Tools:
Whimsical, Adobe Xd, Figma
Team:
Adithya HS, MS Computer Science
Elisa Krebs, MS HCI/d
Harish Maruthi, MS Computer Science
Yuxing Wu, PhD Informatics
Contributions:
Survey, User Interviews, Research Synthesis, UX Design, User Testing, Presentation
Mealtimes are essential and cherished parts of everyday life. On university campuses specifically, public dining areas are positioned to be social spaces to make new connections; however, with the prevalence of digital devices, technology usage introduces a new dynamic into students’ mealtimes.
How might we support and encourage mealtime socialization for college students eating alone on campus?
Our Assumptions
Yuxing’s PhD research focused on mindful eating at mealtimes.
Because of this, our goal had originally been to design and build a socially engaging educational tool to help students make their mealtime more mindful and satisfying.
Looking back, choosing a direction this early on in the process was extremely flawed and our team had to make several assumptions in order to move forward.
Discovering the Truth
Our original project goal and assumptions were leading the project in a very specific direction. Since she was the Project Lead, I respected Yuxing’s previous research in the field and her aspirations to design an intervention. But as the User Research Lead, I advocated for a more generic research approach to make sure we were really solving the right problem and designing the right solution.
The three main research questions we aimed to answer were:
What are the current eating and technology usage patterns of students?
Do students’ technology usage patterns change based on context?
How do college students feel about using technology during mealtimes?
After conducting an initial literature review, we conducted 3 user studies.
A Tough Pill to Swallow
The insights that began to emerge from each research study threatened to upend our entire project. Our focus had centered around mindful eating at mealtimes, but as we sat with the data, we struggled to think of a single viable, user-centered solution in this design space.
We realized that we were being held back by our initial previous goals and assumptions. In fact, almost every assumption was proved wrong in our user research. We needed to pivot.
We combed back through the research and found that the most common theme was social interaction. Not only did social interaction reportedly lead to more enjoyable mealtimes, but it also led to less technology usage. We reframed our How Might We and moved forward in a new direction.
The Pieces Fit Together
Each research method yielded its own valuable insights and informed the way we structured the studies that followed. After analyzing the data and synthesizing our findings, 5 key insights emerged.
Ideating with Engineers?
As the only designer on the team, facilitating 2 ideation sessions for non-designers was a transformative experience for 2 reasons.
This was my first time ideating possible design directions with the folks that would actually be building the designs.
This was my engineers' first time ever participating in a formal design process.
We all had to work hard to change our ways of thinking. For Yuxing and me, most of our ideas were not feasible given the short timeline for development. For Harish and Adithya, knowing the technological constraints made it hard for them to push into dreaming up crazy ideas.
In the end, we ended up with 15 interesting and worthwhile design directions to move forward with.
Whiteboard remnants of brainstorming session 2.
There Can Only Be One
To narrow down a design to pursue, we analyzed each of our concept ideas using a diagrammatic semantic differential. We used a 2-factor model to evaluate our ideas based on feasibility and impact.
We ranked each concept’s feasibility in terms of:
Technical feasibility
Feasibility of actually promoting social interaction at mealtimes if implemented
We ranked each concept’s impact in terms of:
Its impact on our target users based on our needs assessment document
Its ability to impact social interaction at mealtimes
We ultimately decided that none of the concepts were strong enough to stand alone, so the most ideal solution would be a combination of the 3 individual ideas.
Prototyping the Experience
Before designing any screens or interactions, we conducted a low-fidelity experience prototype with potential users.
Participants sat at a 2-person table and were asked to press a button if they were willing to engage in conversation with a stranger. Once both participants pushed their respective buttons, the first conversation question appeared. If a participant pressed ‘Next’ or ‘Skip’ we presented them with the next question.
After 5 questions, we asked follow-up questions and concluded the experiment.
“At some point, I think I’d get to the point where we don’t need the questions to keep going.”
“There’s a lot of times where I’m sitting there and I’ll kind of wish that someone would sit down with me. I wonder if people actually would want to chat or if they’re just being polite.”
Testing Takeaways
The 2 prototype sessions offered us valuable feedback, but also left us with questions.
We wrestled with questions surrounding gender differences and cultural norms in eating environments.
We wondered if students would feel completely different about the interaction if other students were watching.
Unfortunately, due to time constraints, we didn’t have time for further testing.
To make sure our users’ feedback was represented, we mapped each major takeaway from our research sessions to a corresponding design consideration for our final design.
Let’s Make Something!
Our three concepts came together in a solution called MealChat. MealChat combines a physical button and an LED table device with two stationary tablets, each tablet facing its own respective seat.
If a student is sitting alone, the light signifies to others in the area that they are willing to share a meal. The tablet only moves on to the next screen when both buttons have been pressed and both LED’s are lit.
Designing for Meals
As our project deadline quickly approached, we were forced to do design and development simultaneously. This required us to work in a very iterative and collaborative way. We conducted design critique sessions for feedback on the UX design and started building the backend database framework.
For the interaction design, I was inspired by the effectiveness of our extremely simplistic experience prototype materials. Such simplicity diverted attention away from the technology while simultaneously redirecting it to the conversation at hand.
Low-Fidelity Screen Sketches
The question bank database holds thousands of questions, but we limited the users to seeing 5 questions before adding in a buffer, based on how we saw conversation flow in the user testing.
Mid-Fidelity Screen MockUps / Happy Flow
Pops of teal make the final tablet UI mockups feel playful and fun, while keeping a simple style and clean flow.
High-Fidelity Screen Mockups
Design Flow
The diagram below shows the interaction design flow for the entire system; both the tablet application and the physical buttons on the table top button device. Interactions with the tablet are intentionally limited to one click, since users might be eating with their hands.
MealChat Means More
Rethinking the role of technology in mealtimes, Meal Chat seeks to support social interaction, rather than hindering it.
We believe this design goes beyond our original, more basic goals to address larger societal goals by removing the stigma around sharing a meal with a stranger. Our design is an attempt to minimize awkwardness by publicly signaling to others that a user is open to sharing conversation during their meal.
This design also aims to make public spaces more about human connection than just coexisting.
Sample rendering of a Mealchat tablet station.
Our working MVP on display at the Luddy School of Informatics, Computing and Engineering showcase.