Gabriella Harari, PhD, an assistant professor of communication at Stanford University, went to college to study psychology at precisely the right time.
It was around 2006, and there was this new thing called Facebook. While Harari was busy learning about identity and personality theory, she also found herself transfixed by the identity revolution taking place all around her on mobile phones and computer screens.
“I would watch how friends would present themselves to me face-to-face in their daily lives and then how they would present themselves on Facebook or Myspace—or how they would be glued to their phone, a device that hadn’t existed a year before,” says Harari. “And I began to wonder, what are the effects of this? We have no idea where this is going to go but it’s going to be fascinating.”
It didn’t take long to set her educational sights on studying these revolutionary platforms. When she entered grad school in 2011, however, few personality research programs were established that examined the intersections of psychology and technology.
She was lucky enough to work with a top personality psychologist, Sam Gosling, PhD, at the University of Texas, Austin, who was pioneering research on how personality is expressed environmentally, in face-to-face interaction – and in the new digital commons.
The newness of these topics presented serious technological challenges and opportunities, and Harari soon became part of a new generation of social scientists trying to figure out how to harness mobile platforms in methodologically sound ways to develop more advanced personality research methods.
“I got into the smartphone sensing area of research because I was looking for a methodological solution for being able to capture real-world objective behavioral data—so that we could use that in complement with self-report methods and survey methods,” Harari says.
“Traditional self-reporting and survey methods can be effective at capturing subjects’ perceptions and feelings about things,” she notes, “but they often don’t reflect objective behaviors that can be captured with passive data collected on smartphone sensors, such as physical movement, social communication, even sleep.”
Early on, it was a field largely populated by computer scientists who were exploring practical applications for data provided by these embedded sensors—including GPS, accelerometers, microphones, cameras and gyroscopes—for potential applications in fields such as health care, transportation, social networks and safety.
Harari was on the frontlines of one of the first studies that used smartphone data collection to measure and analyze behavioropens in new window (PDF).
In the summer of 2013, Harari visited the lab of Andrew Campbell, a computer science professor at Dartmouth College, to develop a cross-disciplinary partnership. She worked with a team of computer science researchers who were using active and passive smartphone data to understand student well-being and lifestyle patterns over a 10-week period.
It was a fascinating confluence of disciplines, she says. “We had weekly lab meetings where they explained basic things to me about computer science: system architecture and how to set up a back-end server. Then I would present the Intro to Psych perspective. It became a useful way to introduce each other’s framework and technology, so we wouldn’t be talking past each other.”
As interest in what is now known as media psychology or the psychology of technology soared, Harari began to be approached by other social scientists who were eager to employ new technologies in their research.
“When I would talk to psychologists or academic researchers at conferences they would say, ‘That sounds amazing, but how do I get started?’ There is a very large methodological hurdle to overcome if you are a social scientist who doesn’t have any programming background and now want to use an app to collect data.
“You have a few options: Hire a commercial platform, use an open-source app that researchers at another institution developed, or get a programmer or collaborator on your team who has those technical skills,” Harari says.
Harari decided to take everything she had learned over the course of her research and make it easier for social scientists who are new to the field to wrap their heads around the possibilities – and pitfalls – of these new methodologies.
In 2017 she published “Using Smartphones to Collect Behavioral Data in Psychological Science: Opportunities, Practical Considerations, and Challenges.”opens in new window The paper, published in Perspectives on Psychological Science, unpacks the powerful sensing and computational capabilities of smartphones. It provides a description of the types of embedded sensors in smartphones and the data sources they collect – and describes the behaviors they are used to measure.
“It was borne out of a desire to make it easier for other social scientists so that this becomes a common methodology. Then we can really start to push our understanding of behavior much further,” she says.
Lately, what has really captured Harari’s imagination is something even harder to track than Facebook usage or texting practices. She has begun wondering: What are the actual benefits for participants who agree to give researchers their personal data?
“Obviously, we have benefits from the science perspective because we can understand real-world behavior and see how it maps to mood, for example,” Harari says. “But I wanted to close that loop and give something to the user – to begin looking at self-tracking as a benefit itself.”
What exactly is the benefit for participants in smartphone-sensing studies who agree to give out very personal tallies of their daily goings-on? And what motivates people on their own to download applications that track and monitor their behaviors, moods and activities?
Harari wanted to examine the possibility of tying smartphone-sensing data collection studies to the motivations of and benefits to target populations.
To test this out, she conducted a study with a group of students to try to determine what motivations would most likely entice them to participate in a self-tracking study.
Interestingly, the study data showed that the motivations for self-tracking that were most frequently endorsed by the students were to help them “improve academic performance, manage time and to-do lists, track exercise or diet patterns, and understand when they are most productive,” according to the study.
Another finding showed that students who reported having higher anxiety or depressive symptoms were more interested in tracking their mood. “It suggests that there is a potential benefit to matching participants’ motivation with the actual study design, maybe tailoring the questions people receive based on their own interests.”
But does this boutiquing of study design create murky methodology that might skew the validity of results?
“If someone wanted to generalize to a broader population they wouldn’t use that approach,” Harari emphasizes. “There is a concern that if you design studies that way, then you are essentially having your participants self-select into your studies. But in my own work it fits in with what I do, so I like the idea.
“One of the things I try to do more broadly is to see if we can figure out ways that we can promote self-insight based on people’s digital data. We’re creating these digital footprints all the time, so why can’t we visualize this and start to understand ourselves better?
“It’s basically impossible for us to be off the grid now. It’s what drives me about this work and why I feel there are broader, more societal reasons why it’s so important. How are we going to handle this?” she asks emphatically, adding: “I’d like to have work that can begin to speak to those questions and also start talking to people, ‘Are YOU okay with this as an individual?’”
Harari recently founded the Media & Personality Lab at Stanford University, where her group conducts research examining how personality is expressed in the physical and digital environments of everyday life.
Gabriella Harari, PhD, is part of a new generation harnessing mobile platforms in methodologically sound ways to develop more advanced personality research methods.
