MIT Media Lab, E14-633
This thesis designs and carries out a series of studies, which collect multi-modal data from wearable sensors and mobile phones with trait data such as personality types, for learning about behaviors and traits that impact human health and wellbeing.
First, Sano conducted several studies showing how multi-modal wearable sensors can improve characterization of sleep/wake states over motion-sensing alone, and characterizing wrist electrodermal activity (EDA) patterns during sleep. We found that EDA helps discriminate when there is improved sleep-related memory consolidation.
Next, with colleagues at MIT and Brigham and Women’s hospital, Sano designed and carried out the first four semesters of the “MIT College Sleep” study, which measured over 103500 hours of multi-sensor and smartphone use data from 166 college students, recruited with their social groups. Each student contributed intensive multi-modal ambulatory data (physiological, behavioral, environmental, and social) for 30 days. Each also filled out standardized questionnaires on mental health, personality, stress, social interactions, sleep and GPA, and provided a measure of dim light melatonin, enabling circadian phase to be measured.
Sano and her colleagues have found many things. Over the 30 days, most students’ mental health scores dropped. The main factor associated with this drop was self-reported stress. They found that more frequent phone users (measuring “screen on” time) had significantly lower GPA and students with low mental health scores used their phones late at night more frequently.
Other findings include data showing that late night light exposure was associated with lower morning happiness. More frequent phone usage late at night was associated with scoring worse on the Pittsburgh Sleep Quality Index. Irregular sleepers showed more negative wellbeing outcomes (high sluggishness, low happiness) than regular sleepers. The average duration of sleep was 7.0 hours. Their data showed that sleep irregularity was more associated with negative wellbeing than sleep duration. Greater self-reported happiness was explained by lower stress, higher conscientiousness, shorter academic activity total hours, and more frequent positive and infrequent negative social interactions.
They also identified behavioral features that explain monthly averaged and daily stress, happiness and mental health drop using wearable sensors and mobile phones.
Host/Chair: Rosalind W. Picard
Cesar Hidalgo, Charles A. Czeisler