Mindful
Overview
About
A wearable device and companion mobile app to monitor stress levels and connect users to mental health professionals. The device detects stress using a bespoke wearable device and mobile app. The system provides users with tools such as stress level monitors, smart suggestions, and professional outlets to reduce stress and its side effects.
My role
This is a group project. I worked with two other teammates. My role focused on:
Generative Research
Primary Research (Survey, interview)
Data Analysis
Product Strategy
Health Persona Creation
Branding
Affinity Diagram
Wireframing
Prototyping (low-fidelity prototype, high-fidelity prototype)
Usability Testing
Design Iteration
Duration
3.5 month project
Sponsored Project (Cognizant)
The Process
Research and Problem Definition
Why we care?
Research indicates that mental stress, and the challenge of coping with it, remains an issue among the Georgia Tech student body. With the potential to later manifest into physical ailments, this health issue is worth addressing early. Therefore, we seek to identify mental health awareness levels among Georgia Tech students and create a method to help them cope with stress. To do this, we must address their current mental state, alter their management styles, and encourage students toward healthier lifestyles while also removing stigma.
The problem
Extreme mental stress has become a critical issue most students have to deal with. There are 180,000 people in the U.K who die each year from some form of stress-related illness (Meridian Stress Management Consultancy in the U.K.). Stress also accounts for about 75% of all doctor visits, where physical complaints include: headaches, back pain, heart problems, upset stomach, stomach ulcers, sleep problems, tiredness, and accidents (Centers for Disease Control and Prevention of the United States). Furthermore, 90% of all visits to primary care physicians are for stress-related complaints (Occupational Health and Safety news and the National Council).
After investigating how stress affects the general population, we also found links between mental and physical health. Notably, several diseases are known to be greatly influenced by stress, including Gastrointestinal diseases caused by excessive stomach acid, including peptic ulcers, ulcerative colitis, and stomach ulcers. Stress may also hasten the development of ulcers by altering the mucous coating that lines the stomach. Moreover, there is strong evidence that psychological stressors represent a significant risk factor for coronary heart disease. Type II diabetes, adult obesity, and breast cancer are also a few of the many physical ailments being reexamined through the lens of being stress-related.
When analyzing the data, we observed that students at Georgia Tech (GT) experience mental stress, and a considerable percentage find it difficult to cope with stress. A majority of participants try different methods for coping, but only a small percentage seek professional help.
It’s crucial to be aware of your mental stress level constantly, but we have noticed that students tend to ignore signs of mental stress in the early stages due to their high workload. Stress level fluctuations are highly variable between students. Since many are still young, it’s essential to understand the mental health of students and take the necessary actions as early as possible. GT is a high-ranking university with a diverse student body, so we decided to develop our design for GT students due to accessibility to that population.
Source:
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3341916/#b7-mjms-15-4-009
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3341916/
Focal Points
Identifying the causes of stress in GT student populations
Investigating current coping methods and strategies
Deciphering research findings and exploring potential opportunities
Empathize
OBSERVATION
We first performed the desktop research and then conducted real-world research. First, we tried to understand the physical appearance of mental stress in public, focused on the GT campus, and did several unobtrusive observations. The observation happened at busy locations like the GT Library, Tech Green, Student Center, and Graduate Studios. We observed students' body language, interaction, and stress expression in public.
Unfortunately, mental health may not clearly manifest in physical displays, and observation alone might not be the best research tool to gain a breadth of insight. So, this remains one of our challenges when conducting these research methods.
There were several shared behaviors were seen across our independent observations:
Cradling head in hand with one elbow on a desk
Stroking facial hair
Head, ear, and arm itching
Stroking facial hair
Bouncing a single leg
Appearance of lack of focus and staring into nothing
Body stretching
SURVEY
After observation, we conducted an online survey to collect a greater range of data from a broad student demographic. We received over 70 responses within 4 days. We sent the survey to all GT students on different platforms: GT Grad Facebook group, Slack channels, and undergraduate classes, etc. We informed participants that the survey takes 7 to 10 minutes to complete and notified them of complete identity anonymity. If they wished to move on to the survey, students were asked to consent. Otherwise, they can choose to opt-out and leave the survey page.
The first part of the survey was designed to get demographic information about the participants, and the second part was designed to get data about the level of experience responders had with mental stress, their reactions to stress, and their actions to mitigate mental stress. The initial questions asked if the GT students experience mental stress, and 100% of the participants answered yes. For the next question, 100% of participants believed that it is important to continuously manage mental stress. In the following question, 65% responded that they find it difficult to get over stress.
In the later questions, the participants were asked if they sought any professional help, and only 33% sought help. Of those, 7% mentioned using medication for stress. When questioned about the causes of mental stress, 28% voted for schoolwork, 20% voted for future careers, and 18% for current work.
When analyzing the survey results, we observed that all the students at Georgia Tech experience mental stress and a considerable percentage find it difficult to cope with stress. A majority of participants try different methods for coping, but only a small percentage seek professional help. For further studies, the success rate of the coping mechanisms that students use and the reasons professional help is failing to reach students should be studied.
INTERVIEW
Objective: Identify themes/ patterns
Participants: 5 mixed-student demographics
Timing: 30-45 minutes
Methods: In-person and remote
We conducted a user interview to investigate the pain points further. We wanted to learn how stress affects the student population, their frustration levels, and their needs. We recruited 5 participants from the screener survey who agreed to provide their contact information for participation in the user interview.
Challenges:
Preventing participant apprehension due to subject
Keeping the interview on point while still allowing participants room to answer, elaborate, and share
Affinity Diagram
Objective: Extract data from the interviews
Sort: Arrange by different themes and relationships
Propose: Suggest high-level information solutions
To best analyze and visualize the interview themes and patterns, we created an Affinity Map which categorized the interview responses. To do so, we transcribed highlighted areas from the interviews onto sticky notes. These statements were written from a first-person perspective, and as mentioned in the interview analysis section, we primarily focused on the facts, emotions, beliefs, and (re)actions divulged by our interviewees.
Through group consensus, we first loosely ordered the sticky notes around general themes found within each blurb. Following that, we placed the notes on the board in an orderly fashion. If the volume of sticky notes under one category overwhelmed the board and if subtopics became apparent, we subdivided groups and tried to follow the Affinity Map methods learned. Succeeding the arrangements, we labeled each category’s theme, and since we actively verbalized and discussed our thoughts throughout the earlier stages, this came rather easily. Although still several steps away from ideating, we were able to pair hypothetical, high-level solutions for each group. We paired them as seen below:
Challenges:
Judging content relevance and level of importance
Differentiating into sub-groups and identifying each primary focus
AFFINITY MAP | DIGITALIZED
DATA ANALYSIS
At this phase, the data we analyzed represented high-level solutions
My Feelings Toward Mental Health -> Raise Awareness and Remove Stigma
My Symptoms Of Stress -> Identify and Become Aware Of Symptoms Earlier
My Personal Solution To Avoid Getting Stressed -> Personalize Therapy Approaches
Realizing The Importance Of Mental Health -> Validate Or Reject Personal Solutions
My Preferences For Stress Management -> Potential Stress Management System & Emergency Stress Help Services
My Communication Approach and Preference -> Student-Centric Communication Platform
My Living Arrangements -> Provide Alternatives
I’m Stressed When I Think About My Future -> Redirect and Minimize Focus
My (personal) Strategies To Cope With Stress -> Clarify Proper Stress Tools & Provide Recommended Strategies
My Academic Studies Cause Stress -> Task Management, Time Management, & Class Workload Restructuring
My Self-Awareness -> Improve Self-Awareness
My Attempt To Cope With Stress (Using Tools) -> Clarify Proper Stress Tools
Key Findings
While our data collection and analysis remains an ongoing process, the current Observations, Survey, Interviews, and Affinity Map point to several main realizations. They are as follows:
School work-related stresses are most common for students
Stress coping methods and strategies are highly irregular
Although stress awareness seems high, few seek help or continue stress management over time
Ideation Matrix
We ideated through a morphological matrix and examined the goals and ideas within our initial brainstorming session. Then, we identified some key factors that add value to our ideation process. Each color line represents of the key facets within our underlying concepts.
Design Direction
Design 1: Wrist-Oriented
We thought wrist-placed products were intuitive. App and mobile devices and features are increasingly familiar. As children, we often learn through games and activity interests, and we think this practice can also be applied toward adult learning.
Design 2: Wearable Apparel
Our thinking here is that people already integrate clothing into a daily routine. We believe that apparel could have a higher function in communicating situational issues, detecting problems, motivating through feedback, and integrating massage therapies.
Design 3: Headgear
Less universal but also a common body location for wearables. Hats, crowns, and headgear offer comfort and also represent personality projections. Taking this idea literally, stress levels will be projected to, i.e., professionals offering care.
Health Personas
We defined two target user types after generating the interview themes and design direction
MAIN CHARACTERISTICS
Design Requirement
Functional Requirements
Monitor stress levels (Must include)
A method for long-distance friends to send some gestures (Could include)
User must be able to use voice input/command (Must include)
Tools to manage stress (Absolutely must include)
Suggest activities/ methods to cope with stress (Should include)
Rapid destresser (Could include)
Non-functional Requirements
Community (Could exclude)
Not stigmatizing (Must include)
Improve communication/interaction with friends (Could include)
Suggesting activities and tips from experts based on users' preferences (Absolutely must include)
How to keep stress in different lives separate (Could include)
Directly connect with mentors/ experts in real-time (Could include)
Mood check-in (Must include)
Interpersonal connections - buddy system (Could include)
Building goals (Should include)
A method for long-distance friends to send some gestures (Could include)
Self-esteem booster (Could include)
Refined Design Direction
Monitoring stress level and mood check-in
Sending gestures and participating in suggested meditation activities
Connect and interact with friends/peers/ experts->setting up goals
CONCEPTUAL MODEL
INTERACTION TYPES
Sending gestures
Connect and interact with friends/ mentors
Mood check-in and send motivation notes
INTERFACES TYPES
Mobile app
Wearable | Interactive
MAPPING
Co-Design Workshop
Co-designing Wearable Technologies with End-users
Goals
The workshop was designed to harness several key insights from the users. Preferred placement for a wearable device, the physical appearance, the type of the device, preferred materials and spontaneous ideas related to positive expression were collected.
Participant: 6 participants
Location: Design Studio
Workshop Activities
The participatory workshop was designed to harness several key insights from the users. Preferred placement for a wearable device, the physical appearance, the type of the device, preferred materials and spontaneous ideas related to positive expression were collected.
The initial activity introduced the participants to the workshop’s intent and outlined the study's objective: to design a wearable device to reduce mental stress. Participants were provided a printout with a body image and were asked to use markers to indicate their wearable device body placement preferences. Three marker colors were given and ranked by favorability. During the latter stage of this activity, participants were asked to share their wearable ideas through sketching and clay modeling.
For the second activity, participants were given a printout with 16 different wearable options and asked to select three preferred options. To identify their selections, a color coding method was used. If they felt a wearable option was missing, the printout indicated an area to sketch an additional product.
During the third task, participants were provided with two books containing 3500 stickers combined. They were asked to select 10 or more stickers that hold the most positive associations to them, and then, they were asked to write a word that represents why they selected that sticker. If preferred, the participants had the opportunity to combine stickers and create a single expression.
For the final task, participants were given a set of materials and asked to rank them based on their preference for a skin-touching wearable. The given materials included synthetic materials, plastics, leather, fabric, etc.
KEY TAKEAWAYS
Smaller products were preferred
Wrists, necklines, fingers, upper arms, ears, and waistlines were the most agreeable locations
Participants were excited and cheerful throughout the process
Participants preferred sketching, sticker choices and word associates to express ideas
Themes around food/cooking, nature/escape, and friends/family were most common
They were apt to share their ideas and explain their thought processes
Main Findings
Wrist location is most desirable
Users prefer personalized products
Users would like wearable location versatility
Technology Research
EDA is a physiological phenomenon that refers to electrical variations occurring on the surface of the skin due to changes in sweat secretion
(Pope & Ryan, 2019)
EDA is one of the best and most widely used discriminative signals, along with the heart rate signal for measuring stress
(Can et al.,2019)
Detect the stress level using an EDA sensor and react based on the stress level
Stress Level
None
Low
Medium
High
Critical
Response
Custom color pattern
Yellow light
Amber light
Orange light
Red light & vibration
Prototyping
Physical Device
SKETCHES and INITIAL PROTOTYPING
FINAL DESIGN IN 3D MODEL
Mobile App
SKETCHES
LOW-FIDELITY PROTOTYPE
MEDIUM-PROTOTYPE
HIGH-FIDELITY PROTOTYPE
Usability Test
EVALUATION GOAL
To understand how real users interact with the product and make changes based on the results.
Identify usability problems
Collect quantitative and qualitative data
Determine participant satisfaction with the product
Our final design involved a physical wearable device and a mobile app connected to the wearable. With usability testing is expected to collect data on user satisfaction with the wearable device and the interactivity of the mobile app.
ACTIVITIES
Checking comfortability of wearing and taking off the device
Confirming the previously collected data on alternative wearing positions
Collecting user feedback on the behavior of the device
Confirming previously collected data on materials and collecting data on color preferences
At this stage, we did not have the mobile app connected to the wearable, and we wanted to investigate the learnability and comprehensibility of the app. We had four tasks for usability testing on the mobile app.
Setting up processes, including setting up a theme and goal and finding an expert.
Checking the previous week's stress level fluctuation
Connect and interact with a close friend
Find a specialist and schedule an appointment at a specific time.
PARTICIPANTS
We recruited 3 participants (25-30 years old) from Georgia Tech to validate our design. This group represented our target audience (potential users.) These participants were tech savvy, meaning somewhat familiar with wearable devices, owning a mobile phone, or an expert in computer experience. This way, participants have experience using mobile apps and are familiar with new software technologies. We tried to recruit a mixed-gender group for a greater range of ideas and opinions.
METHODS
Several methods were used to collect and record data.
Video recording/ Photographing
All the tests were video recorded with the consent of the participants. Photographs were taken where necessary for documentation purposes.
Think out-loud method
A think-out-loud method was used throughout the test, and users were encouraged to give feedback anytime during a demonstration or when completing a task.
Observation
One team member was placed to observe the participants' behavior and their feedback throughout the tests.
Questionnaire
A questionnaire was given at the end of the test to assess their satisfaction with the wearable and the mobile app.
PROCEDURE
PHYSICAL PROTOTYPE
Task 1
Put on the wearable device on your wrist.
This task is measured using observations. We were looking for signs of struggle and difficulty putting on the device, and these actions were measured.
Task 2
Take off the wearable device and replace the band with the provided extra band.
This task is measured using observations. We were looking for signs of struggle and difficulty putting on the device, and these actions were measured.
Since the prototype was only a medium-fidelity prototype, we used Augmented reality to show how the prototype and each color would look on their wrists. Participants could experience many colors and materials.
DIGITAL PROTOTYPE
Task 01
Complete the setup process
Choose the theme ‘Ocean’
Setup goal as ‘Build Social Skills’
Set the good time for focus as ‘Randomly during the day’
Select an expert “ Social Skills Psychologist.”
In this task, the number of errors made and the number of times the participant got stuck were measured.
Task 02
Check the last week's stress level fluctuation
Go to Home Screen
Go to the full report.
Go to the mood check page and check last week's stress level fluctuation.
In this task, the number of errors made and the number of times the participant got stuck were measured.
Tasks 03
Check the last message from Trish
Go to the home page
Go to my profile and connect with Trish
Check the last message from Trish
In this task, the number of errors made and the number of times the participant got stuck were measured.
Tasks 04
From the home page, go to the professional help page
Find a Trauma specialist
Setup an appointment with a trauma therapist (Lindsay) expert for the earliest available time on the 28 Monday between 9 am-12 pm
In this task, the number of errors made and the number of times the participant got stuck were measured.
QUESTIONNAIRE
Complete the given questionnaire
In this task, the participant had to answer 12 questions in total, and from them, 5 were rating questions, and 7 were open-ended questions. The answer sheets were assessed in a quantitative method.
APPARATUS—the physical setup for the study
The working circuit was developed using Arduino and a potentiometer was used to mimic the signal output from the EDA sensor. A prototype was created using 3D printing and the vibration motor and LED was connected to the prototype using wire.
A leather strap was used as the body connector since it was one of the favorite materials for the wearable among co-design workshop participants.
RESULTS
FINDINGS
On the whole, participants responded very well to our prototype, and their responses further backed up our earlier research findings. After the process of discovering, defining, ideating, and designing around an HCI approach to student stress management, it was exciting to test and validate whether our methodology successfully addresses our user group’s needs or not.
From the start of the testing, participants demonstrated a strong understanding of how and why the product functions and they were able to navigate easily through the service end of the product.
With the physical prototype and in regard to body positioning, the wrist area remained the first choice for a wearable location. Ankle and neck areas were also explored, and like our earlier findings, these related to other preferred wearable locations. Interestingly, the choice to have a chest area placement was suggested, and this was not marked as a preferred area during our co-design workshop.
Through the augmented reality product feature, participants could get a sense of body fit and the many surface color and texture choices available. The same features discovered through the interviews, surveys, and workshops, personalization, were similarly desired during testing. Although participants responded very well to the simulated experience, they wanted to see more product form and color variability.
The mobile interface section of the product was most successful for usability. Minor changes were pointed out to ensure design consistency and comprehension, such as typeface size, menu bar visibility, and a legend for the stress level graph.
MOBILE APP
DESIGN ITERATION
Getting help | Making Appointment
Checking the previous stress fluctuation | Connect and chat with a close friend
LIMITATIONS
User feedback presented conflicting information
Not all app features tested
Questionaire prompts possibly caused erroneous responses
Physical prototype limited user experience
The prototype does not include an EDA sensor
The simulated experience did not test participant stress levels while interacting with the product
On the whole, our follow-up questionnaire showed mostly positive feedback. Interestingly, there were some response inconsistencies. This was noted in answers to statement 4: I thought there was too much inconsistency in the product. Participants all selected strongly agree. However, their responses to statements 5 and 6 indicate high user confidence. It is likely wording or inclusion of statement 4 needs to be reconsidered, and this may have caused confusing responses. It is unclear yet if this is the case since statement 2 shows the greatest measure of variability.
That being said, the general limitation of our study came down to participant numbers. Ideally, we would have included more than 3. Our goal was five to ten. If we had been able to secure more participants, our current findings might have been further validated, or we may see the opposite as true. Moreover, more testing would help to present app features that were not presented to participants during this round of testing. Participants wondered why our product did not include a wanted feature when in fact, it does. However, they were not guided through the entire interface during testing and failed to experience the full range of features. If a little more time were available, a design investigation just on the physical prototype would have been valuable.
References
Pope, Gunnar & Halter, Ryan. (2019). Design and Implementation of an Ultra-Low Resource Electrodermal Activity Sensor for Wearable Applications ‡. Sensors. 19. 2450. 10.3390/s19112450.
Can YS, Chalabianloo N, Ekiz D, Ersoy C. Continuous Stress Detection Using Wearable Sensors in Real Life: Algorithmic Programming Contest Case Study. Sensors (Basel). 2019 Apr 18;19(8):1849. doi: 10.3390/s19081849. PMID: 31003456; PMCID: PMC6515276.