Fragile Families study spawns research and teaching opportunities
It was just before noon on a Thursday in late September when a group of professors, graduate students, postdoctoral fellows and staff members assembled in a teleconference room in the basement of Wallace Hall.
As they got settled in, a voice called out from one of the screens at the front of the room. “This is the University of Pennsylvania. Can you hear me?”
“Yes,” came the reply from Sara McLanahan, professor of sociology and public affairs and the director of the Bendheim-Thoman Center for Research on Child Well-being.
Soon students and faculty at Columbia University could be seen taking their seats on another screen.
“So is everybody ready?” McLanahan asked. A few moments later, Julien Teitler, an assistant professor at Columbia’s School of Social Work, began presenting his findings on the effects of welfare participation on marriage.
The teleconference, which meets every other week, brings together
students and faculty at the three universities who are conducting
research using Princeton’s Fragile Families and Child Well-being Study.
The study, now in its fifth year, has gathered reams of data about
the families of 5,000 children in the United States to gain insights
into unwed parents and the effects of parental resources and public
policies on children’s well-being. The collection of data is available
to the public so that researchers everywhere can make use of it. The
study, which released its third wave of data last month, has spawned
dozens of theses and dissertations by students and more than 70
academic papers by scholars at Princeton and at many other
“In terms of basic research, it’s the most extensive data we have on unmarried partnerships with children,” said McLanahan. “People all over the country are using the data for their research, which is exactly what we want to happen.”
The three-way teleconference is one example of the many ways that the Fragile Families study serves not only as a source of critical social science data but also as a valuable teaching tool for students.
‘Great modeling for being an academic’
When Teitler concluded his presentation, he was peppered with questions from participants at Princeton and the University of Pennsylvania, who suggested ways that he could improve his paper, which is a work in progress.
“The teleconference is fabulous for our students,” said Jeanne Brooks-Gunn, a visiting research collaborator at Princeton’s Center for Research on Child Well-being and a developmental psychologist. “They get to see how researchers collaborate, how different authors of a paper answer questions differently and how people in different disciplines interact. It’s great modeling for being an academic.”
Because everyone presenting research during the teleconferences is using data from the Fragile Families study, students are deeply engaged in the material.
“When students attend a workshop, they usually know little about the data on which the research is based, so they have a limited ability to participate,” McLanahan pointed out. “Here students have a chance to contribute, to share tips, make suggestions. Once they’re using the data, they become experts. It gets them talking and feeling competent.”
Sharon Bzostek, a graduate student in sociology, is working with the data to explore the ramifications for children of mothers having romantic relationships with men who are not their fathers.
“Even though I’ve been working with the data for less than a year, doing so has already taught me a tremendous amount about conducting longitudinal data analyses and has introduced me to some specific issues and questions that often arise when studying family dynamics and child well-being,” Bzostek said. “I feel very fortunate to have had these experiences during my first year of graduate school.”
The teleconference has become an integral element of a
graduate-level course being taught in the University of Pennsylvania’s
sociology department by associate professor Kathryn Edin. The yearlong
class is focused on studying the Fragile Families data, and each
student will write a paper using the data that will be submitted for
“The hardest thing for a student is to think of a research problem, to conceptualize an idea,” said Edin, who has been conducting research for the last four years on 75 families from the Fragile Families study. “It’s a great experience for them to hear models of researchable problems during the teleconference.”
Understanding unmarried parents
The Fragile Families study collects information from participating parents, who live in 20 U.S. cities, during hour-long telephone interviews in which hundreds of questions are asked about income, employment, health, housing conditions and attitudes about marriage. The study seeks to understand what the capabilities are of unmarried parents, especially fathers, what the nature of the parents’ relationship is and why they often choose not to marry. Interviews for the baseline data were conducted at the time of the child’s birth. The current data release encompasses interviews with the family when the child was 3 years old.
The baseline data found that most parents were romantically involved at the time of the child’s birth, and two-thirds expected to marry. But the follow-up study found that while 16 percent of the couples did get married, 47 percent were no longer together by the time the child turned 3.
The latest round of data collection included in-person interviews for the first time. Christina Paxson, professor of economics and public affairs at the Woodrow Wilson School, obtained a grant from the National Institute of Child Health and Development which was used to conduct in-home assessments of 2,581 families. The interviewers evaluated the children’s cognitive development, looked for physical and emotional neglect and examined whether the children were living in a safe environment.
“It’s the kind of information you can’t get over the telephone,” Paxson said. “The interviewers are testing the kids and observing the environment.”
The interviewers even brought scales to the homes to weigh parents and children, which led to one of the study’s notable findings: Obesity is apparent in children as young as 3 years old.
The next wave of data collection is going on right now as the children reach age 5. McLanahan has submitted grant proposals to reinterview the families when the children are 9 years old.
“Every year the data gets more valuable because you have more of the
family’s history,” she said. “I wish we could keep it going through
adolescence and adulthood and even into old age.”