Product Designer
compass-intro-image full.jpg

Compass

Improving access to educational video content through inclusive design.

 

Compass

 

The Problem

In times of stress, change, or uncertainty, it's easy for us to become disoriented from our values, emotions, and self-care habits. In this project, I designed for individuals going through a major life change - those who are experiencing a prolonged period of stress, change, or uncertainty.

The Challenge

How might a self-guided utility help users understand and improve their moods during times or stress, change, or uncertainty?

Solution

My team designed a mood management system to help users keep track of their moods and to provide users with suggestions for mood-improving activities when they need it the most.

My Role

UX Designer; I led the design phase, which included facilitating a brainstorming session, creating paper prototypes, designing wireframes, and creating an interactive prototype. 

Tools: Sketch, InVision

Team: Joey Hoy, Skye Zheng

Timeline: September 2016 - December 2016

Type: Class Project

 

The Approach

 
 

User Research

We began this project with a limited understanding of our target user. We had a number of research questions we wanted to explore:

  • Who are the people who could use more support during major life changes?
  • What kind of support do they look for?
  • When would support be most welcome?
  • How do people currently track their moods?
  • How do people use data from their mood-tracking systems?


To learn more about our users, we employed the following research methodologies:

     

    SURVEY

    By collecting data from a large number of respondents, we were able to obtain a broad overview of our user group. Main findings:

    • 75.5% of survey respondents have experienced a major life change within the past year; 83% of these respondents indicated that they could use more support during major life changes
    • 21.1% of respondents currently track their moods
    • The reported benefits of mood tracking are: reflection, future reference, and viewing the cross-impact of moods and health

     

    USER INTERVIEWS

    To complement our broad data set, we conducted in-depth interviews to explore the goals and contexts of our user group. Main findings:

    • Major life changes are disruptive to routine and self-care
    • Participants are more likely to track their moods during emotional highs & lows
    • Participants rely on constant sources of support, even during times of stability

     

    PARTICIPATORY DESIGN

    We gave participants the tools to imagine their ideal solution for managing moods during times of stress and change. This was also a chance to learn about the benefits and opportunities to improve upon existing products they might use. Main findings:

    • Show empathy/concern in phrasing
    • Provide self-guided resources (e.g., mindfulness exercises
    • Propose coping mechanisms one or two days in advance of emotional peaks and valleys

     

    COMPETITIVE ANALYSIS

    We conducted a competitive analysis by exploring and comparing features of 7 popular and/or participant-mentioned mood tracking applications. Since many of our survey participants indicated that they use life-tracking or mood tracking tools, a competitive analysis of current offerings helped us to better understand the benefits and areas of opportunity.

     
     Artifacts from two different user interview sessions

    Artifacts from two different user interview sessions

     
     

    Defining a User-Centered POV

    Our research led us to develop the following personas. Personas promote empathy and help us to focus on user needs (attitudes, behaviors, motivations, and frustrations) during the design process

     
     
     
     

    Defining Core Functionalities

    We determined that the core functionality for Compass would be to suggest individualized self-care tips to users during times of emotional stress and uncertainty.

    This was determined based on our survey and interview data. Our survey revealed that 76.1% of respondents had experienced a recent major life change or disruption to routine (within the past year), and 83% of these respondents indicated that they could use more help caring for themselves during these times of change. Our interviews revealed that participants were more likely to reflect upon their moods during emotional highs and lows, whereas times of emotional stability did not encourage tracking or reflection.

    Based on these findings, we designed Compass to be used during these times of emotional extremes.

     
     
     

    Design Requirements

    • Allow users to take their time and process/be mindful in the moment of entering because that in itself is useful practice
    • Propose coping mechanisms or actions a day or two in advance
    • Show empathy/concern in phrasing
    • Use facts for suggestions that indicate it might work for you rather than condescending or presenting tips as though our system knows better than users do
    • Allow users to receive and act on notifications without interrupting their routine if they don’t want to configure that
    • Support higher-level organization across days and lower-level structure within each day / entry
    • Provides self-guided resources
    • Provide an option to log on a weekly or monthly basis, or whenever the user chooses
    • Provide encouragement
    • Support pre-defined/user-generated tags or icons that allow for easy input
    • Include context and metadata to accompany each entry
     
     
     

    High-Level Process Flow

     
     
     

    Ideation

    During the ideation phase, we explored a variety of ways to bring our design requirements to life. We completed individual sketches, then came together as a team to determine which designs we wanted to build and test.

     
     Early versions of our design explored a chatbot feature to facilitate mood inputs and to provide suggestions through a friendly and encouraging tone of voice

    Early versions of our design explored a chatbot feature to facilitate mood inputs and to provide suggestions through a friendly and encouraging tone of voice

     
     

    Low-Fidelity Paper Prototypes

    We created paper prototypes for our first round of user testing and evaluation. The low-fidelity nature of our prototypes invited user feedback and allowed for quick iteration between testing sessions.

     
    Paper Prototypes.png
     
     

    Mid-Fidelity Clickable Prototype

    Following our first round of user testing, we incorporated the changes into an interactive prototype, created using Sketch and InVision.

    View the prototype here: https://invis.io/CHA1HT2FU

     

    Onboarding

     
     
     

    Mood Input

     
     
     

    Receiving Suggestions

     
     
     

    Evaluating Usability 

    We completed two rounds of usability testing - the first using paper prototypes, and the second using a clickable prototype. We asked participants to complete three tasks while we observed their interaction with the system:

    1. Onboarding
    2. Mood data input
    3. Receiving suggestions and providing feedback

    With each round of testing, we produced a list of design recommendations that would help improve the usability and clarity of our design. Our key findings include: 

    • Provide access to saved moods through chart and calendar views
    • Add scroll indicators to guide users through information
    • Improve upon conversational component via word choice
    • Reduce the amount of text presented during the on-boarding process


    We tested our paper prototype with 6 participants, and 45 usability issues were discovered and reworked. With our interactive prototype, we tested 3 participants and addressed 34 usability issues.

     
     
     

    Reflection

    For me, this project was an interesting exploration into predictive AI and its applications for mental health. If we had more time, it would be interesting to test how suggested task size (e.g., small suggestions, larger interventions) would influence user engagement and completion rates. With more user data and the application of machine learning, it may be possible to create models that help us understand how mood patterns map to certain task preferences, and whether these models reflect larger user psychologies. 

    Finally, I think it’s important to continue asking ourselves: should mood management remain a non-technical process? While there’s no clear or definitive answer to this, I do believe that it’s important to revisit this question when designing technical interventions that replace historically non-technical approaches.