StudentLife: Paving the Way for More Ubiquitous Mental Health Solutions
Looking at the evolution of technology, especially cell phones we have come a long way into ubiquitous computing in the past few decades. We have gone from simple cell phones that was just used for calls and messages to smartphones that use the internet and can allow a user to do a variety of things such as email, make payments, make schedules, track a person’s steps, and more. There are quite a few renowned institutions making remarkable contributions to ubicomp. One leader in ubicomp research is the Dartmouth Networking and Ubiquitous Systems Laboratory (DartNets Lab). They are making contributions to track and help with student mental health by implementing Smartphone Sensing.
The DartNets Lab is a fairly new lab founded by Dartmouth College’s Computer Science Faculty in September 2013, focusing its research on ubicomp applications and networking of smart devices. The team is directed by Professor Andrew T. Campbell, Professor Temiloluwa O. Prioleau, and Professor Xia Zhou and composed of post-docs, graduate students, and undergraduate researchers at Dartmouth College. This lab has published many papers and constantly had their work highlighted at the annual ACM International Conferences for UbiComp (Ubiquitous Computing), MobiCom (Mobile Computing and Networking), MobiSys (Mobile Systems, Applications, and Services), and more.
One of the lab’s research areas is Smartphone Sensing. This involves using sensors from smartphones to track user behaviors to assist with mental health studies. The DartNets Lab’s first major project was StudentLife which is still used for ongoing studies. StudentLife is an app that uses smartphone automatic sensing to track and monitor students’ behaviors for the StudentLife Study at Dartmouth College. The StudentLife Study began in the Spring term of 2013, tracking 48 Dartmouth students over the course of 10 weeks.
The StudentLife app utilizes computational methods and machine learning algorithms to assess the smartphone sensors to make high level inferences of human behavior such as sleep, stress, sociability, physical activity, and other activities. The smartphone automatic sensing allows human behavior to be captured in the background, so no user input is needed to obtain the info. The purpose of the StudentLife app is to assess the mental health of students based on their activities as well as its correlation with academic performance. The results of the study showed a common trend among these students which is termed “the dart-rhythm”. By midterm season, physical activity significantly declines while mood and conversation duration slightly decline and remains steady until the end of the term. Stress levels go up during midterm season and remain steady until the term finishes.
The StudentLife Study got nominated for the Best Paper Award at the ACM UbiComp’14 and was discussed at the conference. This study then got media coverage from a variety of outlets such as BBC World News, CBS News, Forbes, and more. The StudentLife team plans on taking this app to the next level and developing StudentLife 2.0. The proposed addition to StudentLife is to implement feedback and intervention into the app. Some examples of intervention involve informing students of risky behaviors such as excessive partying, social isolation, and poor sleep. The goal is to promote healthy living, safety, and improved academics around campus.
Building off of the StudentLife Study, DartNets Lab developed the SmartGPA Study based off the StudentLife app. Similarly, smartphone automatic sensing monitors a student’s behaviors as it did in the StudentLife Study. However, this study focused on comparing the habits and behaviors of low academic performers and high academic performers based off of the StudentLife dataset and finding the correlation of student behaviors and GPA. The analysis from this study has resulted in the development of a model that predicts a student’s GPA within ±0.179 points of their actual GPA, based on their behaviors. The SmartGPA tool is a proposed tool to predict a student’s GPA on the model developed by the study. The SmartGPA paper received an Honorable Mention Award at UbiComp’15.
The research done with StudentLife and SmartGPA paves way for the future of mental health resources. The population of the StudentLife study can be expanded to more students at Dartmouth as well as more colleges to gain more data to better define the GPA prediction model from SmartGPA. Expanding these studies can allow for such an app proposed in StudentLife 2.0 to be developed on a national level for the use of all college students. Adding the intervention aspect that suggests students to steer away from health and academic threatening behaviors can overall help the mental and physical well-being of the student population as well as promote healthy and productive lifestyles.
The future of Smartphone Sensing research can help more with mental health in general as well as physical health. Apps can be developed similar to StudentLife to track a person’s behaviors and provide early diagnosis of common mental health problems, indicate if a user is at risk of developing an issue, inform the user on the indicated signs, provide suggestions on improvement of their behaviors, determine how strongly in need the user may be of professional help. The same can be applied to common physical health problems on a smartwatch app with sensors that sense heartbeat, physical activity, etc.
After the 2020 pandemic, many people developed mental health issues such as depression and have had dilemmas with access to mental health resources and social interaction. This situation has shown the need for mental health resources being available on ubiquitous platforms, particularly laptops and smartphones. Currently, there are apps such as 7 Cups which allows an individual to anonymously share their concerns to a listener and Calm that helps with sleep and meditation using relaxing music and stories. However, these apps require user input and interaction with the app.
There are people in need of some type of help who may be unsure of what kind of platform is helpful for them, so they don’t end up downloading anything. With something similar to the StudentLife 2.0 proposal, a user’s habits can be tracked over a period of time and feedback can be provided to the user to inform them of their risky habits and suggest the proper kind of help. For example, a user with inconsistent sleep patterns can receive a notification that they need to improve their sleep and be provided suggestions to improve sleep. Another example, a user’s sociability may be low which is impacting their mood and stress so the app can inform them of this and provide appropriate suggestions.
Using the StudentLife projects as a backbone, there can be significant progress made in ubiquitous mental health solutions in the upcoming years. The awareness of the subject of mental health has had a lot of growth in the past few decades and is gradually receiving more significance in the eyes of ubicomp research. However, with developments of apps and technologies to assist with mental health, there needs to be appropriate marketing and advertising of these apps through various outlets as well as being promoted in schools so more people can have these apps on their devices. With a combined help of ubicomp research and social awareness of mental health, improvements can be made to benefit the people of our society.
