GIST Spring 2013 by Duke SSRI

The social science magazine of Duke University Spring 2013, VOLUME 7, ISSUE 1

gist

Working

together

to change the lives of children and families

Nechyba’s Niche

Questions:

Tom Nechyba sits down and discusses the latest on SSRI West here as an infrastructure resource in multiple dimensions, all of which are directly relevant to the Bass mission. The Information Initiative Duke (iiD) and SSRI’s Data Information Services will both be housed in Gross Hall in an effort to help researchers work with their large datasets. How would you describe “big data” and how can these two groups help researchers? Nechyba: “Big data” can mean lots of different things—everything from very large traditional datasets to data derived from audio, video or text coming through old and new media. In some cases it means data on lots of people; in other cases it means data with lots of dimensions even if the sample itself is small. It is in this sense that the old distinction between “qualitative” and “quantitative” social science breaks down. For instance, rich ethnographic data on a small number of families may in fact be “big”, just as data on social networks from text messaging or Twitter data are big. The work of Guillermo Sapiro and Helen Egger (more on page 2) applies “big data” methods to an initially small sample of a few hundred videos, but then holds the promise of developing applications that can be applied in truly groundbreaking new ways around the world. Then there are data derived from virtual environments—such as those generated by the work of Jeff Taekman (more on page 4). And, as previously disconnected data sets like those contained within our Census Data Center are merged, these data too can become “big” pretty quickly. The time really is ripe for productive new collaborations between social scientists and “big data” methods to emerge, and having SSRI in the same space as the iiD is very promising.

Social and behavioral scientists often work together on research projects, but SSRI West plans to be the home of “research teams” with faculty and other researchers from various disciplines working together. What types of projects would you like to see held at SSRI West? Nechyba: The best collaborative projects for SSRI West will ultimately not be the ones I envision, but will be driven by the interests of faculty and students who engage with the space and the infrastructure resources we offer. That is the vision for SSRI West— that it gives the intellectual, physical and infrastructure space for innovative ideas to emerge in collaborations that otherwise would be difficult to conceive and carry out. But there are things (other than providing infrastructure and cookies) we can do to facilitate collaborations. For instance, together with the iiD and our Duke Network Analysis Center, we hope to use information on what faculty are writing about and proposing in grant applications, to create an intellectual map of unexplored potential faculty networks that may then become actual collaborations in SSRI West. SSRI West will be home to research teams instead of actual affiliates and programs— the way SSRI is currently structured. Do you anticipate these teams being long running or once the problem is solved, a new team project being created? Nechyba: We live in a world that is changing more and more rapidly, and I think it is important for universities to find ways of becoming more nimble. Not everything worth doing right now has to become a permanent center that takes on a life of its own, and so I hope SSRI West can be an incubator of projects that will evolve and change and yes, even end. Sometimes a problem will be solved,

Duke University Photography

Bass Connections, a new university-wide initiative launched by a $50 million gift from Anne and Robert Bass, was recently announced. What role will this initiative play with SSRI West? Nechyba: SSRI West is almost ideally suited for supporting a variety of Bass Connections projects that intersect with the social sciences. The layout provides physical space for teams to engage and then to linger in subgroups within what we are calling The Connection— a large open and welcoming area for spontaneous meetings and engagements. At the center of The Connection—right next to the cookies and coffee—is the help desk we are calling The Connection Bar. Any Bass Connections group that relies on social science methods and data support— everything from ethnographic work to big data computation—will find The Connection Bar, as well as the various workshops offered in The Connection very useful. We are also planning data and methods “bootcamps” specifically targeted to meet the needs of research and project teams, essentially providing backbone support to enable students at all levels to become highly productive team members. So we are

gist Spring 2013 vo lu m e 7 , i s s u e 2 1

other times progress will have been made giving rise to new projects that can push the frontier. SSRI currently offers workshops and a help desk but the space at SSRI West will enable education and training for social and behavioral scientists to be revamped. We’ve heard about the possibility of an app as well as a staffed help desk with “experts” in various software and data programs. Who do you envision utilizing these services? Nechyba: The problem right now is that we try to bring many of our workshops to West Campus but, since we have no permanent location there, they move around and it’s hard to keep track. Our help desk is at our off-campus location, and so few students or faculty take advantage of it. But our SSRI West help desk—The Connection Bar—and the related workshops and “bootcamps” will all be located in the same place near the heart of the West Campus and within a newly renovated building buzzing with activity. We are still working out how exactly it will all function, but yes, we are thinking of a SSRI West app that will enable students and faculty to always know what is available and when, allow appointments to be made from anywhere, and stream online help on common issues. Who do we envision utilizing the services? We’ll have workshops to train undergraduate and graduate RAs who can then rely on The Connection Bar for back-up; projects that draw on The Connection Bar for expertise and advice; faculty who feel more at ease forming interdisciplinary teams with the knowledge that The Connection Bar can assist. And I hope it will be a draw for our affiliated centers, a place to launch new collaborations and connect to new faculty and students with the support of The Connection.

Tom Nechyba Director, SSRI

ssri . d u k e . ed u / gist

2 Working Together To Change the Lives of Children and Families

Training in Virtual Environments

Exploring a Whole New World of Social Data

Gaining Access to Confidential Data Without Compromising Individual Privacy

Marriage & Fertility

Influencing Social Policies Through Data

Bass Connections

Managing Editor: Courtney P. Orning courtney.orning@duke.edu

Scan this QR code to read GIST online.

Contributors: Paul Dudenhefer Whitney Howell Michele Lynn Nancy Oates Mary-Russell Roberson Designer: Regina Barnhill-Bordo regina@bdesign-studio.com This publication is printed with vegetable-based inks on chlorine-free paper containing 10% postconsumer fiber. Please recycle this magazine.

sites.duke.edu/ssriwest/ facebook.com/DukeSSRI twitter.com/ssri youtube.com/ssriutube flickr.com/photos/ssri_duke/

The Social Science Research Institute at Duke University is a part of Interdisciplinary Studies at Duke University

www. s s r i . d u k e . e d u 1

Working Together to change the lives of children and families Fewer than one in four children with an

2 gist • Spring 2013

“Everything from a literature review to automatic video analysis, data analysis, installation in the school and working with the kids as we collect data.” Egger and Sapiro will move their teams to SSRI West, space under renovation in Gross Hall that will enable social scientists and experts in big data and information futures to cross paths and share ideas. Sapiro had recently come to Duke from the University of Minnesota, where his research included analyzing videotapes of a preschool classroom and at a clinic to screen for autism. Egger has been at Duke since her residency and for years has been conducting large-scale epidemiological studies of anxiety disorders in preschoolers. For the past decade in the Duke Early Childhood Research Program, she and another colleague, Dr. Adrian Angold, have assessed more than 2,000 preschool-age children, including videotaping assessments with more than 500 children and their parents.

Clockwise from top: Sapiro and Egger pose with their MRI machine; waiting room for children and families; faculty, researchers and students discuss their project on identifying at-risk children.

BWPW Photography

impairing psychiatric disorder ever comes to the attention of any mental health professional. Yet increasing evidence shows that early intervention helps children develop more normally and improves their functioning as adults. Finding a reliable, affordable, nonintrusive method to identify children who need help is the sort of complex, realworld challenge that excites researchers and demands a multifaceted approach. Two researchers at Duke, from disparate fields, are collaborating to create diagnostic tools that could lead to early identification of at-risk children. Dr. Helen Egger, chief of the Division of Child and Family Mental Health and Developmental Neuroscience at Duke University Medical Center, has several years’ worth of videotapes of children and their parents that must be coded manually, a time-consuming process. Pratt engineering and computer science professor Guillermo Sapiro has the video imaging analysis expertise, as well as proficiency with novel computational approaches to multimodal large data sets, to develop new methods for automated analysis of these data. “You can manually annotate 100 videos,” Sapiro said, “but not 10,000.” Jointly, Egger and Sapiro are putting together a project that unobtrusively installs inexpensive video cameras in a Durham public elementary school classroom and in Duke’s Early Childhood Research Lab to

record anxiety behaviors of children, and then automatically analyze the data. Sapiro envisions giving video equipment to Duke students who travel abroad, enabling him and Egger to do cross-cultural studies examining, for instance, whether the behavioral triggers of anxiety are the same in Tanzania as in the U.S. The results could lead to an inexpensive way to identify children from all walks of life and across cultures who could benefit from support services. “This is not ivory tower research,” Egger said. “We’re doing work that we hope will change the lives of children and families.” Egger and Sapiro are augmenting their research team with the help of Bass Connections, an initiative announced in January funded by a $50 million gift from Duke trustee Anne Bass and her husband, Robert. Bass Connections aims for a new education model that gathers scholars at all levels, from undergraduates to postdocs and professional-school students to senior faculty, across all 10 schools at Duke to tackle interdisciplinary research on realworld problems. Within days of the Bass Connections announcement, Egger and Sapiro had interviewed four students who wanted to join the team. “We have multiple things for students of almost any discipline to do,” Sapiro said.

“ What’s exciting about this collaboration is it’s built on the work I’ve done throughout my career and work he has done in his whole career. It’s a wonderful opportunity to do research in our lab and translate and test it with real kids in their everyday environment.” —Dr. Helen Egger

“We have one of the largest data sets with in-depth, multimodal assessments with community populations of young children and their families,” Egger said. “These data, collected over more than a decade, provide a unique opportunity for data mining.” Developing automated, time-efficient, individualized methods for analysis and interpretation of data opens the door to analysis of big data of magnitudes never studied before in this discipline. The collaboration, enhanced by the Bass Connections team, marks the first time Egger has worked with undergraduates. On the faculty of Duke Medicine, she teaches and mentors medical students, residents, and postdocs, bringing an understanding of the medical education model. “In doing, you learn,” she said. “That’s how you train in medicine.” The interdisciplinary, multilevel team Egger and Sapiro are building epitomizes Duke’s long-standing culture of collaboration across schools and disciplines, said Susan Roth, vice provost for interdisciplinary studies. “We’re building integrated pathways for students who want a particular expertise that intersects with real problems in the world,” Roth said. “Conversations will run

across the university among people who wouldn’t typically meet.” Hallie Knuffman, the liaison from the Office of Interdisciplinary Studies who coordinates Bass Connections, sees Duke at the forefront of such interdisciplinary research. “Collaboration is easier to do at Duke than at other universities,” she said. Even the layout of the physical campus helps. “At a lot of universities, the medical campus is miles away,” she said. “At Duke, it’s a 10-minute walk from the main campus.” Students who join the Egger-Sapiro team have the opportunity to work with engineers, mathematicians, developmental epidemiologists, early childhood psychiatrists, and developmental neuroscientists to apply computational analytic approaches and data mining to the videotapes Egger has compiled already. The team also plans to use robotics to develop cost-effective, easy-to-use tools to collect data. “Everything opens tons of research questions,” Sapiro said. “We don’t know how to solve all these problems. A tough problem is great for my students and for me.” Egger recently received a $1.6 million Duke Endowment grant to build an integrated pediatric mental health coalition

from Duke and the community that includes public schools, pediatricians, and mental health professionals to improve the mental health care for children in Durham. She and Sapiro have applied for additional grants to cover the cost of equipment and to support their study. Within the next few months, Egger and Sapiro hope to install a data-collection videotaping system in a classroom in one of the poorest performing public schools that serves some of the neediest students in Durham. “What’s exciting about this collaboration is it’s built on the work I’ve done throughout my career and work he has done in his whole career,” Egger said. “It’s a wonderful opportunity to do research in our lab and translate and test it with real kids in their everyday environment.” Sapiro finds the immediate application of his research very satisfying. “One of our main challenges as a society is to help our children,” he said. And the collaborative approach benefits researchers, too. “I’m learning a lot,” he said, “because I’m working with people who are very far from my own area of research.”

www. s s r i . d u k e . e d u 3

Training in

virtual environments Better Than Listening to a Lecture

r. Jeff Taekman encourages

4 gist • Spring 2013

In fact, Taekman said he first realized video game technology might be useful educationally when he watched his young children immerse themselves in games for hours. “I got started asking myself the question of why children will spend so many hours playing video games,” Taekman said. “What is that secret sauce that will keep someone in front of computer games for hours when they can’t stand to sit through a lecture for 20 minutes?” As it turns out, that magic ingredient is “flow”—that surreptitious knowledge acquisition that occurs during interactive or entertaining activity just beyond a learner’s current limits. Such an engaging, effective

Photo courtesy of Jeff Taekman

his medical students to play video games in class. In fact, he lets nursing students, residents, and attendings get in on the action too. The only caveat? He gets to record and analyze what they’re doing. These aren’t ordinary video games, though. They’re virtually immersive programs designed to help students and practicing healthcare providers improve their skill sets, in the hopes of augmenting patient care. Each decision—every action— is recorded for detailed study. Traditionally, Taekman, (assistant professor of anesthesiology, assistant dean for educational technology, and the medical director

of the Human Simulation and Patient Safety Center in the Duke School of Medicine), has used patient simulators or mannequins to help students and providers hone their techniques. It’s a good teaching method, he said, but it also has its limitations. “Many years ago, I became interested in the use of video game technology because I was frustrated with the scalability and distributability issues with mannequins. They’re basically controlled plastic dummies,” he said. “But virtual environments, in addition to being infinitely scaleable, generate a lot of data about the choices people make and what they do. There’s incredible potential for analyzing this data and marrying it with clinical indicators or teaching outcomes.”

primary practitioners or ophthalmoloSSRI West, we’re going to be able to underteaching tool “is the Holy Grail of what gists—cognitively work through the thinkstand what works.” we’re shooting for,” he said. ing they’ll have to do when sedating an The future of big data and video gaming These games give participants immeadult or child,” Taekman said. “This will doesn’t solely lie in education and profesdiate feedback and allow them to work become part of a multimodal simulation sional training, however. It also has the through scenarios until they complete them program targeting non-anesthesiologists potential to solve some of science’s most successfully. Depending on the rules of the who provide sedation in the military and in complicated and intriguing problems. For program, time in the scenarios can speed civilian settings (e.g. gastroenterologists example, some games pit players against up or slow down, producing data about for colonoscopy). each other as they fold chains of virtual how providers respond outside of normal amino acids to determine the 3-dimensional constraints. In this context, Taekman said, To date, project results have shown that shapes of proteins. Another game aims data means big data—massive amounts individuals who train in virtual environto improve sequence alignments for of discrete information points generated ments, similar to those who train with by multiple individuals participating in the game or training session. Dr. Ed Buckley (vice dean for education in the School of Medicine), Taekman and others received generous funding from the Duke Endowment to build a shared common virtual learning platform called ILE@D (Immersive Learning Environments @ Duke). The first modules based on this platform are just now coming to fruition. Ultimately, the goal is to improve the overall efficiency of simulation, known to be “ Marrying that data together is going to help inform us as to what a very effective—and expensive—training technique. The collected data will help works and what doesn’t work in education. These games can providers better understand the mix of pump out massive amounts of data on every player, so for the modalities used in simulation, Taekman first time, working with SSRI West, we’re going to be able to said, as well as shed light on where and how virtual environments can play a role in understand what works in education.” positively impacting healthcare quality. —Dr. Jeff Taekman And, the training possibilities are far promoter regions in a plethora of diseasefrom exhausted. mannequins, retain information far better related genes by asking participants to In February, Taekman launched a protothan if they listened to a lecture. shift color-coded blocks, representing DNA type game to study whether the same stratBut having all this data associated with bases, right or left to determine the best egies used with students would be equally training and teaching isn’t worth much match for several species simultaneously. effective with practitioners already in the unless it can be processed and analyzed. clinical environment. In addition, he is That’s where SSRI West comes in. Its unique Existing interactive video games frequentalso using funding from the Telemedicine mixture of interdisciplinary faculty offers a ly used to promote fitness can also be & Advanced Technology Research Center substantial potential benefit to Taekman’s effective in treating the symptoms of within the U.S. Department of Defense, to research. Having all the collaborators he autism, Taekman said. These games engage support research into whether the virtual could possibly need under one roof elimithe mind and the body, and monitoring gaming strategy can effectively train nonnates both the need to seek out partners, as how patients with autism-spectrum anesthesiologist physician reservists the well as the likelihood that he’ll need to ship disorders play, researchers can change cognitive skills needed to provide sedation his data out for analysis. and refine treatments. to soldiers and indigenous populations in “That data is going to help inform us as “There are many novel ways games are combat zones, even if they have no previto what works and what doesn’t work in being used in scientific discovery, and that’s ous experience with anesthesia. education,” he said. “These games can an area we’re interested in expanding into, “We’ve developed a series of 10 cases pump out massive amounts of data on every as well,” Taekman said. “I don’t think of to let these physicians—whether they’re player, so for the first time, working with games as purely being a teaching tool.”

www. s s r i . d u k e . e d u 5

Exploring a whole new world of

social data This summer, a group of statisticians, mathematicians, and engineers will be moving into Gross Hall, one floor up from SSRI West. The group is collaborating to build a Big Data toolbox—techniques to analyze enormous datasets such as the billions of daily posts on social media websites; or the 600 million proton collisions per second at the Large Hadron Collider; or the three billion A, C, G, and T letters in the genomic sequence of a single person. Humankind is creating so much data, at such an astonishing rate, and in such varied forms (numbers, words, images, video) that the standard tools for data analysis are no longer up to the task. “It’s really exciting because there are people from a variety of backgrounds,” said David Dunson, a professor in the department of statistical sciences, who will be moving to Gross Hall with the Big Data group. “I’m a statistician—doing inferences, allowing for uncertainly—but there are also excellent mathematicians, who are really good at characterizing low-dimensional structures and high-dimensional data, and electrical engineers who are really good at algorithms.” In high-dimensional data, each datapoint has many variables—for example, a group of patients (the data points) and billions of pieces of genomic information (the variables) about each of them. “When you have millions and billions of observations on a given patient, and the number of patients

6 gist • Spring 2013

Humankind is creating so much data, at such an astonishing rate, and in such varied forms (numbers, words, images, video) that the standard tools for data analysis are no longer up to the task. you have is much smaller, you can’t analyze that using traditional methods,” Dunson said. Instead, the strategy is to create a “lower-dimensional structure” that will allow the meaning to shine through. Dunson uses Bayesian statistics as a route to low-dimension structure. Others, such as Rebecca Willett, a self-described electrical engineer with a mathematical bent, use geometrical tools to help describe complicated structures in simpler ways. “Think about your sheet in a dryer,” she said. “That sheet is just a flat, two-dimensional surface, but then it’s crumpled up and floating around in a three-dimensional space. It’s no longer flat, but still, if you were an ant, at any one point it looks flat.” This concept—called manifold—can be applied to high-dimensional datasets to create something simpler than can be analyzed. Tools developed to analyze high-dimensional data in one setting, such as astronomy, can be applied to high-dimensional data in another setting, such as the social sciences. “They are pretty different kinds of problems on the surface, but they all have similar types of challenges that show up,” Willett said.

The new Information Initiative Duke (iiD) in Gross Hall will be co-led by Robert Calderbank, dean of the natural sciences in the Trinity College of Arts and Sciences, and Larry Carin, the William H. Younger professor of electrical and computer engineering. Others who will set up shop in the space include Ingrid Daubechies, James B. Duke professor of mathematics; Mauro Maggioni, professor of mathematics and computer science and electrical and computer engineering; and Guillermo Sapiro, professor of computer science and electrical and computer engineering. Willet stated, “We all bring a different set of tools to bear on this big underlying problem, but right now we’re scattered across campus. After the move to Gross Hall, we’re hopeful we’ll be able to accomplish a lot more as a team than we’ve been able to.” The design of the new space—with plenty of common areas and white boards—will encourage interaction, and a regular influx of new ideas and energy will be provided by visiting faculty, both from Duke and from the larger scholarly community. Willett said the move will create new opportunities for students as well. “Right now our students

don’t interact that much, or when they do see each other, it’s a seminar and there’s no time to chat,” Willett said. In Gross Hall, “They’ll be together all the time and will be able to gather around a white board on the spur of the moment. My students will get to learn from the statisticians, and the statisticians will get to learn from the mathematicians.” Being just one floor up from SSRI West will provide the scientists of the iiD plenty

of what they crave: more data. Social scientists have traditionally gathered a lot of data through surveys, but now a whole new world of social data is available. Dunson says an emerging field, which many of his students are interested in, involves writing programs to “scrape” data off web sources, such as Twitter, Facebook, and Google. Willett hopes SSRI will provide not only more data, but also new challenges. “It’s

exciting that we might be able to help out with social problems via these collaborations, and it would be great if we have to develop new tools to look at these problems from an entirely new perspective,” she said. “Not only do I hope to be able to help SSRI, but I hope they’ll be able to help me figure out new and exciting areas of math to study.”

www. s s r i . d u k e . e d u 7

Gaining access to

confidential data without compromising

individual

privacy

To most people, the U.S. Census Bureau is simply the government office that, every ten years, counts the number of people who live in the United States. But to many researchers at Duke University, the Census Bureau is a rich source of demographic and economic data—data that make it possible to answer all sorts of important and interesting questions. A lot of the data, however, are confidential, and rightfully so: would you want any member of the public to have the ability to look up your salary or age? So how are researchers at Duke able to access and use them? Enter the Triangle Census Research Data Center. One of the new occupants of what will open in the fall of 2013 as SSRI West, is the Triangle Census Research Data Center (RDC), one of fifteen Census Data Centers across the country that allow qualified researchers access to otherwise confidential

8 gist • Spring 2013

data. What makes the data so special—and so sensitive—is that they are linked to individual people and businesses. And having access to data on such a “micro” or individual level makes it possible for researchers at Duke to answer questions that would be impossible to answer with the kind of aggregate data (state unemployment rates,

for instance) that the government normally releases to the public at large. One of those researchers is Deborah Rho, a Ph.D. student in economics. Rho is collaborating with Seth Sanders, a professor in the economics department, and director of the Duke Population Research Institute, on a project that studies the earnings of recent immigrants to the United States. To what extent do the earnings of recent immigrants resemble those of native-born workers? Do the earnings of recent immigrants rise as fast as those of the native-born? What distinguishes their study from previous ones is that Rho and Sanders are able to examine not just earnings patterns, but earnings patterns in relation to the kind of firm for which an immigrant works. And such an inquiry is made possible only by the access to confidential data that the RDC provides. “We would not be able to link earnings with workplace characteristics without the work that the Census is doing and without the access they grant me through the RDC,” Rho said. The Census, through the RDCs, provides

15.6%

N u m b e r o f i m m igrants that m a k e u p th e U . S . wo r k f o r c e

rho’s research

17 million

BWPW Photography (Deborah Rho Portrait)

an infrastructure to do research in a way, Rho stresses, that does not compromise the privacy of the data. Another researcher at Duke who is using the RDC is Daniel Xu, a professor of economics. Xu is working with a team of scholars to study the effects of government subsidies intended to increase the number of dentists working in rural, underserved areas of the United States. The problem is more complicated than it may at first seem. In most of those areas, there is already one, sometimes two, dentists, and they are often making good money, since there is so little competition for their services. What happens when a new dentist, with the aid of a government subsidy, moves into the market? The profits of the existing dentists fall as they lose business to the new dentist; as a result one of the existing dentists may move his or her practice elsewhere. The net result? The same number of dentists as before. Confidential data provided through the RDC allow Xu to closely track the movements of dentists into and out of a particular market. Without having to reveal information on individual dentists, Xu is developing a cost-benefit analysis that will determine how much money the government needs to spend in subsidies in order to increase the number of dentists in a given underserved area. “The data are confidential, but the Center should not be a secret,” said the director of the RDC at Duke, Gale Boyd. He says that

N u m b e r o f f o r e ign - b o rn p e o p l e that ha v e b e e n grant e d r e sid e n c y stat u s sin c e 1 9 9 0

people come knocking on the RDC’s door after they have squeezed all they can from the aggregate data the Census makes public. The questions that people really want to answer require data on the level of the individual person and the individual firm. Those are the kinds of data the RDC provides. Boyd’s own research has focused on energy efficiency. He is currently writing a paper with a former student in the master’s program in economics at Duke, Mark Curtis, who is now in the Ph.D. program in economics at Georgia State University (located in Atlanta, a city that also houses a RDC). The two are looking at whether the management practices of firms influence the energy use of those firms. Such an empirical analysis could not be done without the confidential, microlevel business data that the RDC provides. To use the RDC, researchers must be granted Special Sworn Status. Applicants must undergo a medium-level FBI background check and submit a work and residence history, fingerprints, and a sworn affidavit that the applicant will protect the data under penalty of law. As Boyd notes, the federal statistical system is broad and rich, but fragmented. He foresees a day when RDCs function as gateways to other federal data. Already the RDC environment, which promotes access to confidential data, is rubbing off on other departments. According to Boyd, the National Center for Health Statistics now offers more broadly provided access to confidential data. And the Bureau of Labor

Deborah Rho

Statistics now uses the RDC model to provide researchers access to confidential data. In the past, it has been mostly economists who have used the RDC. But Boyd hopes that the move to SSRI West will draw a broader range of social scientists to his Center. In fact, that is already happening, he said. “Because RDCs have historically dealt mostly with business data, economists have been more likely to know about them and use them. But in the last 5–10 years, lots of rich demographic data have been added to RDCs, so the availability and usefulness of RDCs to other social scientists are growing.” Ever since his days as a Ph.D. student in the 1980s, when he began working on energy use in industries, Boyd has recognized the value of microlevel data; and ever since then, he’s worked to make those kinds of data available to more and more researchers. There are often institutional, technical, and legal barriers to making confidential data available, but one by one they are being overcome. “There was a time when you could not get access at all to confidential, microlevel data. Confidential, microlevel business data were opened to the research community—but you had to go to the Census headquarters in Washington to access them. Now you can go to an RDC.”

www. s s r i . d u k e . e d u

marriage & fertility Heather Rackin’s research focuses on an institution that profoundly shapes society—the family. Concentrating on the family and fertility, she is interested in how inequality results in different family forms. Rackin is currently exploring the links between marriage and fertility with a focus on how cognitive schemas and material conditions interact for low-income Blacks. Rackin, PARISS Fellow (Program for Advanced Research in the Social Sciences), and post doc in the department of sociology, uses both qualitative and quantitative methodologies to achieve her research goals. With hours and hours of in-depth interviews from the field, she has chosen to use Network Text Analysis (NTA), combined with traditional qualitative techniques, to work with her data. Network Text Analysis is a method for encoding the relationships between words in a text and constructing a network of the linked words as a way to pull meaning out of text. “If two words are in the same paragraph together, it probably means they are linked together cognitively. If one word brings up another, it means they are more closely connected in people’s brain and are more likely to be associated. You let the computer do it for you. The hard part isn’t the programming— it’s understanding what it all means,” explained Rackin. 10 gist • Spring 2013

Rackin warns that the methods aren’t going to give you the question. “Everybody goes into research with certain underlying questions and assumptions. NTA can be overwhelming with the amount of data you get back so you need to know beforehand what you are asking. If you don’t have an idea of your hypothesis before you run NTA, you better have a high tolerance for frustration.” As part of the PARISS Fellowship yearlong program at SSRI, Rackin is part of an elite group of graduate students with an annual stipend, office space, computing, and other research and administrative support. All fellows are advanced graduate students working toward their dissertation, with interest and achievement in quantitative social science. Rackin feels the PARISS fellows program works so well because the group is cross-disciplinary. She believes graduate students are most likely to try new things and push past the standard methodologies, but it can be difficult to get credibility, funding, and the support needed to pursue the research. Although the fellows are engaged in various research topics, they are able to share ideas, methodologies, and offer advice to one another. When describing the PARISS program, Rackin explains, “They understand what you’re going through when you become

Heather Rackin conducts ethnographic studies examining the links between marriage and fertility with a focus on how cognitive schemas and material conditions interact for low-income Blacks.

BWPW Photography (heather rackin portrait)

frustrated or stuck on the research. We’re all doing cool and innovative research and we’re able to meet up and talk it out.” said Rackin. Rackin is trying to answer the question, “What are the ideas of low-income Blacks before they have children and what kind of schemas do they have that surround children or marriage?” Her project is important because in the current research there is a perception that the idea of marriage and child bearing are two disconnected decisions with different prerequisites—this has different implications for policy. Contrary to previous findings that low-income parents do not link marriage and fertility and have different requirements for each, Rackin has found that marriage and childbearing are indeed linked and have similar requirements prior to childbearing. Low-income Blacks hold quite traditional views about the appropriate role of marriage and its sequencing vis-à-vis fertility. She argues that the material constraints to marital childbearing may lead to non-marital births and therefore respondents sever schemas connecting marriage and childbearing and adopt other schemas of childbearing to provide ad hoc justifications for their behavior. Much of the previous research was conducted on unwed parents. “It’s hard to say that what we find in lowincome unwed parents applies globally to low-income individuals and communities. Aspirations for middle class are ideal, but being able to get it is a different story. And

if you can’t get it, you may change PARISS and SSRI leadership about ways the the way you think about things,” program may be reshaped or augmented to explained Rackin. recruit graduate students into the PARISS program at an earlier stage in their studRackin has always been interies. The program will link these new PARISS ested in the topic of family. As an affiliates with research teams such as those undergraduate she conducted an developed through Bass Connections, SSRI evolutionary biology thesis on affiliates, and individual labs. The students’ fertility patterns and sexuality in monkeys. When she arrived at Duke roles on these teams will both advance the teams’ research agendas and enhance the she was able to collaborate with students’ training through intensive trainfaculty working on similar research ing in methods coupled with a genuinely projects including S. Philip Morgan, previous research-centered interdisciplinary year. The director of SSRI, and Christina Gibson-Davis, expectation is that students who participate associate professor of public policy. Much of in this new element of the PARISS program the data Rackin uses is from Gibson Davis’ will, thanks to their training and team-based projects. After learning about Jim Moody’s accomplishments, have a better chance Duke Network Analysis Center (DNAC) and securing PARISS Fellowships to advance its methodologies on social networks, she their dissertation research. became interested in the network research. She has learned how to adapt Moody’s methodologies for words, instead of the more s t u dy r e s u lt s traditional data, to better understand how people’s cognition influences their behavior. Rackin agrees SSRI West is a great move for the PARISS fellows. In order to be successful fellows, students need Married couples the opportunity to collaborate with faculty and other graduate students. Even just talking Number of women out a research problem over Number of times a cup of coffee during an specific words came impromptu conversation in up in transcripts The Connection could yield great results. Number of men Number of people in study* Rackin’s ultimate methodological goal is to mesh Number of times child quantitative and qualitawas mentioned tive analysis to create a hybrid Number of times marriage and child were mentioned that includes the strengths of both methodologies. She hopes 2,707 she can help bridge the qualitative/ Number of times marriage was mentioned quantitative divide while enlarging and transforming the scope of possible answerable questions. *Each participant of the study was Black, in lowincome communities, unwed and had no children With SSRI’s planned expansion to SSRI West, talks are underway between the

17,562

3,897

899

www. s s r i . d u k e . e d u

Influencing Social Policies

Through Data Linda Burton cares deeply about children and families living

12 gist • Spring 2013

intensity of data,” said Burton. “We have lots of detailed information about individuals.” She says that while in many ethnographies, it is common to find studies with a sample size of 40 to 50 families, the big science ethnographic studies, with which she is involved, goes far beyond that. Burton’s research falls between being qualitative and quantitative. “The gathering strategies we use are qualitative interviews and participant observation; we write field notes and transcribe interviews,” she said. “But all of that information can be turned into numbers. Calling it qualitative or quantitative is a false dichotomy in some ways with this work.” She considers her work to be ‘big science’ not only because of the large size of the ethnography itself, but also because of the reach of the work. “We are able to connect to other types of methodologies,” Burton said. “You can do integrative papers with survey data and other kinds of data sets. For example, I’m really hoping to see links created between ethnographic research and research in neuroscience.” SSRI leadership is excited about collaborating with Burton and her team, which will bring important methodology into the new space at Gross Hall. Burton expects SSRI West to facilitate links between diverse researchers. “I’m hoping that the space will allow lots more conversations, not only about substantive issues, but also about the methodological challenges in interpreting both qualitative and quantitative data and the mix of those types of data,” she said. Burton has found that some of the most significant breakthroughs she has had with data analysis, or thoughts about her research families, have occurred in organic conversations with colleagues. “And that’s what I’m hoping for in this new environment,” she said.

Photo courtesy of Linda Burton

in poverty. That passion has inspired Burton, the James B. Duke Professor of Sociology, to focus her research on these individuals and the impact of public policy on their lives. “I want to make sure that the voices of these families are heard in both the policy realm and the realm of scholarship in ways that can actually help influence social policies to enhance their life situations,” she said. As principal investigator of the ethnography component of Welfare, Children and Families: A Three-City Study, Burton focused on the impact of welfare reform on families and children. Developing a better understanding of how growing up in rural areas might influence the development of children and their families was the focus of her work with The Family Life Project. Burton characterizes her research projects as ‘big science’ studies, which typically involve researchers from a number of different disciplines. “I work with economists, anthropologists, other sociologists and those who work in the areas of human development and social work,” she said. “We come together to figure out ways to gather the best kind of data and to gather that data using different techniques.” Utilizing a mix of research strategies allows scholars to discover things that might go undetected when using only one research approach. “While surveys, for example, can find broad sweeping themes with respect to certain phenomena about poverty, my research—which is ethnographic in orientation—provides the opportunity to see the more nuanced processes that go on in people’s everyday lives,” Burton explained. “The ethnography helps us interpret the broader themes that survey data might discover in greater detail.” “When I think about big data in ethnography, I think about the

Bass Connections Interdisciplinary Learning. Real-World Issues.

In January, the university announced the creation of the Bass Connections initiative, named after Robert and Anne Bass who have generously pledged a $50 million foundational gift to launch it. But three months later, many faculty and students are still wondering: What is Bass Connections? It’s a good question, and there are good reasons why it’s not easy to answer. But at an abstract level, here is one way to think of it: Over the past few decades, three pillars have come to be viewed as central to Duke’s identity as an institution: • Excellence in Education— undergraduate, graduate and professional, • Cross-Disciplinary Engagement— in problem areas that transcend disciplines

(and most challenging) aspects of Bass Connections center around the idea of bringing the three pillars together in the form of project teams that engage problems in the world. These teams will focus on problems of societal importance (Pillar 3) that, by their nature, require disciplines to come together (Pillar 2) while engaging learners at all levels (Pillar 1). They are vertically integrated in the sense of engaging undergraduates, graduate and professional students, post-docs and faculty; and they are horizontally integrated in that they engage multiple disciplines within the university as well as (potentially) representatives from outside academics. Initially, Bass Connections and the accompanying Bass Connections Teams will be piloted around five themes that

have strategic priority for Duke: (1) Global Health, (2) Energy, (3) Information, Society and Culture, (4) Brain and Society, and (5) Education and Human Development. In the long run, however, the initiative is envisioned to expand beyond these themes, with teams connected to and integrated with other activities in the research and teaching space at Duke. For an increasing subset of the faculty, this could provide an opportunity to more fully align teaching and research while at the same time orienting student involvement away from a purely traditional course-based curriculum and causing departments to re-think what it means to develop disciplinary expertise and to contribute to solving problems. And places like SSRI will play a critical role in developing the infrastructure to support the teams that lie at the core of Bass Connections and that epitomize what interdisciplinarity at Duke is about: the coming together of strong disciplines to solve problems that require more than one type of expertise. Additional information about Bass Connections can be found at: www.interdisciplinary.duke.edu/

• Knowledge in the Service of Society— with a focus on problems of societal importance

There are, of course, many ways in which these pillars already intersect at Duke. But unlike previous initiatives, Bass Connections attempts to explicitly and substantially connect all three pillars while boldly inviting these connections to cross all schools, institutes, and initiatives throughout the university. It therefore isn’t so much another new initiative—rather it is a framework for connecting the various elements of Duke that have grown up all around us. And that is precisely why it has been hard to articulate what “it” is. What “it” is will invariably emerge rather than be defined up front—“it” will be shaped by those who engage with and help forge the web of connections that is Bass Connections. While some of what will emerge will come in familiar form—new courses, speaker series, workshops— the most unique

bass C o nn e c ti o ns

E xcellence in Education

Cross-Disciplinary Engagement

Knowledge in the Service of Society

www. s s r i . d u k e . e d u 1 3

Non-Profit Org U.S. Postage PAID Durham, NC Permit No. 60

P. O . B o x 9 0 4 2 0 | D u r h a m , NC 2 7 7 0 8

gist c e

n i o t e c

Mingle

Step into The Connection

vending

Cookies k i tc h e n

Writable Glass Wall

Connection Bar

Coffee

Connect

Team Projects

with SSRI West

Study or Chat

Collaborate with other researchers. Meet up in our team rooms. Grab a cup of coffee and mingle. Share ideas on writable glass walls. This is where ideas happen! sites.duke.edu/ssriwest/