As of 2017, an estimated 17.9 million adults in the U.S. suffered at least one major depressive episode.

Source: National Institute of Mental Health

Depression and anxiety affect millions of people around the world. One helpful and commonly used coping mechanism is journaling. Keeping a written record of one's emotions and thoughts can provide valuable insight into our internal worlds. However, quantifying trends over time can be quite difficult when your primary source of data is natural language. Alternatives to journaling such as spreadsheets with numeric scales correlated to various emotions offer a way to both record data and visualize trends. However, this method poses a moderate barrier to entry due to the effort required to build the spreadsheet, input data, and generate visualizations.

Design Brief

Our approach to solving this problem was to build a mobile application that gave individuals an opportunity to collect self-reported data about their mental state, paired with graphs illustrating trends over time. We encouraged the user to record their mental state regularly through daily reminders, an easy way to input data, and readable visualizations.

Outcome

As a result of our efforts, Lume Wellness helped a number of users record and better understand changes in their mental states over time. At the peak of Lume's popularity, we maintained an active community of over 2,000 active users per month.

Case Study

Though mobile-based alternatives to spreadsheets existed, many featured complex and/or confusing self-reporting interactions. The data visualizations also offered limited value in many cases, either by being hard to read or by failing to provide insight into time-based data. Additionally, few personal health trackers allowed the user to compare their mood data with other types of information such as diet, activity levels, and location. Our hypothesis was that by simplifying the act of self-reporting and delivering meaningful and readable visualizations, we could help individuals gain a deeper insight into their mental health.

To validate our idea, we interviewed people we identified as potential users of the product. This included family, friends, and extended contacts. We also gathered responses to a short survey distributed among a variety of online communities. Through our research, we were able to determine that 39% of all respondents already practiced some form of journaling, while 64% the same group stated that they would use a product that offered an easy way to self-report and visualize trends in emotional states.

How do you quantify emotion?

Our first steps were focused on solving two problems which formed the core interactions of the user experience: devising a simple method of tracking emotion, while visualizing the data in a way that highlighted trends in mood over time. We also wanted to provide additional context for the mood data, such as location, day, time, and a short text entry (a little longer than a tweet).

To begin, we needed to develop a better understanding of emotion itself. What is emotion? What are its constituent parts? How does it affect our bodies? Our perception? Conversely, how do factors both internal and external influence our emotions?

Next, it was necessary to determine how such information would be captured and quantified. How does the user record their mood and energy? This was one of the toughest interaction problems we faced. Do we present an array of sliders and dials, allowing for very precise and detailed input? Or do we simplify reporting emotion to selecting a single value? Each approach had it’s own set of benefits and drawbacks.

Trying to understand how to map emotions.

Design through iteration.

Early iterations simplified the experience, guiding the user through recording their mood, and then their energy. We devised spectra made to be unambiguous in language and assigned an odd-numbered scale to each, with the values going from 1 to 7 (Panas scale). Precisely defining what we mean when referring to emotion or mood played a key role in designing the interactions for self-reporting.

We worked to design a fun and easy interaction for reporting, eventually utilizing a circular number dial for the input of mood and energy data. Generous tap targets on the circular dial and a large number/text display made input a quick and easy affair.

Two core principles guided the visual design from the beginning: simplify the complex, and be friendly. Data visualizations that were easy to find and read formed the core of the design. We worked to strike a balance between a minimal interface that wouldn't distract from the data, but would still feel friendly and inviting.

Visual Design

Early iterations of the product centered around the idea of a friendly AI who lived on your phone. We named them P.I.P. - short for Personal Information Pal. We wanted the visual design of the interface to look forward, exploring the future of mobile design. Despite this, care was also taken to not stray too far into unfamiliar territory, which could potentially alienate users. A minimal approach to the visual design allowed the data visualizations to be the focus.

DIN Pro Rounded was used for both its effectiveness at small sizes, and the distinctive nature of its letter forms. In the context of Lume’s design, the type face provided a modern - but friendly - aesthetic. Avenir Next was the typeface of choice for rendering journal entries and non-numerical copy.