MIT professor Rosalind Picard is worried about campus stress.
After a handful of suicides in recent years, Picard started thinking about how her own work might be able to help change MIT’s emotional climate. The founder and director of the Affective Computing Research Group at the university’s Media Lab, Picard has long tried to turn emotions into something that can be counted and measured — following the “old engineering principle that for something to be controllable, it has to be observable.”
Simply asking people about their stress levels doesn’t work. “Usually they’ll say, ‘Oh, I’m fine,’ ” said Picard, who is also faculty chair of the school’s MindHandHeart Initiative, a campus-wide effort toward student wellness. People may not even know themselves when they’re becoming more vulnerable.
For the last five semesters, she and Akane Sano, who just received her doctorate in affective computing and is now a research scientist in Picard’s lab, have been tracking student stress in one of the largest studies of its kind.
In their five 30-day studies, they’ve followed up to 50 students at a time with stress monitors, twice-daily questionnaires, cellphone trackers, nutrition check-ins, and some physiologic measures of signals like melatonin levels.
“We have been using machine learning techniques to find out . . . the significant difference in behaviors between high-stress and low-stress people,” Sano said.
One finding of their studies is that students with poor sleep quality were more likely to be highly stressed. The researchers also found an association between low mental health and high stress. And they found that students who talked on their phones late at night were more anxious than those who didn’t.
The studies are too short-term to prove what caused what, Sano said. Did students make phone calls because they were stressed, or did those late-night calls add to their tension?
Picard and Sano are now planning a yearlong study, hoping it will enable them to predict oncoming problems, like depression.
But for people to stick with a study like that, it has to be fun. She’s been playing around with inserting goofy questions along with the serious ones. “Can you multiply?” the trial program recently asked the MIT professor. “How much giraffe are you?”
Once she’s compiled the data, Picard hopes the evidence will be convincing. People who don’t sleep much and show signs they’re on a downward trajectory, might be shown data that those who sleep more are less stressed, mentally healthier, more creative, and have higher GPAs.
“I’m not going to tell people what to do,” she said, but then asked, with a smile: “What time are you going to bed tonight?”