New Professor Uses Cell Phones to Gather Data on Teen Stressors
With her palpable enthusiasm and warm smile, Candice Odgers, one of the Sanford School’s newest faculty members, gives you the sense right away that she is accessible. So it’s not surprising to learn that collaboration and innovation are two key aspects of her research. Odgers came from UC, Irvine, to Duke this summer as an associate professor of public policy at Sanford, associate professor of psychology and neuroscience, and associate director of the Center for Child and Family Policy.
She and her research team are using technology in ways relatively new to social science research. To build a fuller picture of the context in which children and adolescents grow up, they are linking up with partners ranging from Google engineers to Verizon Wireless project managers.
In her miLife study, for which she partnered with Verizon Wireless, Odgers and her team distributed free cell phones to adolescents in California so they could track daily happenings at a key stage of life that holds clues to future substance abuse, mental health problems and obesity risk.
In another project, Odgers and her researchers are using Google Street View to take a virtual walk down London streets and record both positive and negative features of local neighborhoods that are believed to influence children’s health. They link this information to an already comprehensive dataset gathered on families in the Environmental Risk (E-Risk) Longitudinal Study, directed by Duke professors Avshalom Caspi and Terrie Moffitt, to create a “genes-2-geography” map of the factors that may influence children’s life outcomes.
Charting her own course
Odgers grew up with her four siblings in a remote village in Saskatchewan, Canada, where she remembers ice skating on the roads and winters so cold that uncovered skin would quickly freeze. She traces her interest in the effects of environment on child development to her work as a youth advocate in the courtrooms and streets of Vancouver, where the harsh realities of growing up in poor and violence-ridden communities and the role of these experiences in shaping children’s life chances could not be ignored.
“Working in the juvenile justice system… you are trying to connect these kids with services and intervene, but in many ways you just wish you had gotten there sooner.”
While Odgers was working one summer to organize a NATO advanced research workshop in Poland, an American professor (her soon-to-be PhD advisor) convinced Odgers to leave her career path in sociology and become a psychologist. She decided to forego a prestigious research award, the Commonwealth Scholarship, and relocate to the University of Virginia to complete her PhD in psychology. Her journey then took her to England for postdoctoral training, before circling back to the United States where she started her first faculty position in California.
“As I moved across contexts,” Odgers said, “I began to understand how the commonalities across disciplines and the fundamental differences between countries in their response to children’s behavior and mental health could be leveraged to better understand and promote healthy development.”
Pinpointing when to intervene to help early adolescents avoid future substance abuse was a key part of the miLife study. Young adolescents were recruited into the study if they had a history of behavior problems or a family history that increased their chances for developing a substance abuse problem. The study was designed to identify the daily triggers of a number of health-risk behaviors and, eventually, inform how and when to intervene in the lives of young adolescents.
Unlimited Texting, Lots of Data
The miLife study’s use of cell phones to collect real-time data is seen as cutting edge in social science. Odgers and her team launched the study in 2009, loaning phones with unlimited texting to about 150 participants aged 13-15. Following a baseline assessment of teens and their parents, the adolescents responded to three short surveys per day for 30 days and reported on their daily interactions and activities. Questions dealt with mood, drugs, violence and experiences during the day. The content of their texts to friends and family also became part of the dataset.
The aim of the miLife study is to determine whether and how the repeated stressors that teens experience across a short period of time accumulate to alter their trajectories during critical periods. The micro-focus was meant to complement the standard child development research, much of which follows children over long periods of time, with months and often years between assessments, and tends to focus on the impact of major life events.
“There is good evidence to suggest that the daily, chronic, pervasive and sometimes lower-level stressors and exposures that many of us experience have the largest effects on people’s health,” said Odgers. “In the same way that the number of words a young child hears in a day differs among families and predicts later verbal ability, the daily stressors, harsh exchanges and uplifts that kids encounter each day are likely to move kids in both positive and negative ways and, in some cases, leave a lasting mark.”
The miLife study provides insight into young adolescents’ lives through an almost instantaneous delivery of information.
“This type of study design gives you the ability to adapt midstream,” Odgers explained. “You have the data coming in, you can see what’s happening—who is responding, who is not. And you have the opportunity to calibrate questions͟, and eventually micro-interventions, in a more individualized way.”
The miLife surveys had an impressive response rate of over 90% and generated approximately 800,000 lines of data, which is still being analyzed. The teens sent as many as 400 texts per day and averaged 100 per day.
Using Google to Assess Neighborhoods
Through support from Google, Odgers was able to fold an additional layer into a very thorough longitudinal study being conducted by Caspi and Moffitt. Over a decade ago, their Harvard-based collaborator, Robert Sampson, tried to help overcome self-reported bias in neighborhood assessments by purchasing a van and having his team drive down the streets where the families in his Chicago-based longitudinal study lived. But because of high costs, only 20 percent of the families’ neighborhoods could be coded. With the advent of Google Street View, Odgers was able to adapt Sampson’s method and virtually collect similar information on the neighborhoods of children in the E-Risk Study for a fraction of the cost.
Demonstrating how Google Street View could provide a reliable and cost-effective way to assess neighborhoods attracted so many inquiries from other researchers, for-profit entities and others that Odgers’ team decided to make the assessment tool freely available online. They hope this methodological innovation will help researchers better understand how the communities where people live, work and play influence behavior and health.
Touching Kids’ Lives
While she may have a history of taking risks, what led Odgers to Duke was a much more calculated decision. She saw a strong link between the mission and focus of the Duke Center for Child and Family Policy and her belief that science should have a real and positive impact on the lives of children and adolescents.
“I was excited about the prospect of being in a place with a true commitment to interdisciplinary research and student engagement, while also having access to the platform that the Center provides to translate knowledge in ways that can reach practitioners and touch kids’ lives.”