GIST Spring 2008

The social science magazine of Duke University Sponsored by the Social Science Research Institute spring 2008, volume 2, issue 1

gist f ro m t h e M i l l

the making of an election exploring demographic issues where the environment and health meet Building a culture of interdisciplinarity at Duke

greetings

third issue of GIST with excitement over the continuing developments at SSRI as it enters yet another stage in its growth. SSRI is now one of the seven signature interdisciplinary institutes at Duke. As such, it is in the process of gaining approval to make research faculty appointments.

Also, SSRI is participating in the Provost’s initiative to hire faculty jointly between departments and institutes. Our plan is that these hires will contribute to the proposed interdisciplinary social science statistics center located within the Institute.

As we grow, we also take stock of where we’ve been. SSRI was founded on a core principle: we support research and teaching efforts across all of the social and behavioral sciences.

This breadth is evident in the research methods we support, including behavioral and physiological assessments in our stateof-the-art experimental laboratory, storage and analysis of large-scale data sets in our new data core, and computer-based questionnaire construction and analysis. It’s also clear from the variety of scholars in our Monday evening speaker series (come join us!). Most recently, speakers have included noted social scientists from outside Duke such as Larry Hedges, education researcher and meta-analyst; Ken Bollen, sociologist; Ken Shotts, political scientist; and Chris Edmond, economist (note that the last two teach in business schools), among others.

This breadth is critical, we believe, to the vitality of the Institute. If it supported a narrower focus, the Institute could not meet its mission of catalyzing the development of new, integrative approaches to issues in the social and behavioral sciences, and could not serve its broad constituents across the University.

Through our founding principle, SSRI provides primarily public and collective goods. It does not value data based on how much they cost to produce or whether the

data align with external funding priorities.

Some research is expensive and requires external grant funding to conduct, whereas other research is relatively cheap. Our grants office helps social and behavioral scientists apply for and manage grant support as needed. But we recognize that the national funding landscape is uneven and is not organized by the contribution and potential of the various disciplines.

Some of our affiliated centers have strong grant support; others have very little. We offer comparable services to all by supporting innovative approaches wherever they arise, regardless of the likelihood of funding. To do otherwise, we believe, would be shortsighted. (Research funding is much like the stock market, with hot sectors that have access to funding changing over time as politicians become interested in new questions.) A broad portfolio of funded and unfunded research activities positions SSRI to identify innovations across disciplines.

In this way, we remain at the forefront not only in the current funding climate, but also when funding priorities inevitably shift.

Given our commitment to support teaching and research across the social and behavioral sciences, we measure SSRI’s success as much by what people learn here as by what they produce here. We are, at heart, an educational institute. This means that we offer many of our services free of charge, in order to promote training of new social and behavioral scientists and retooling of ones already in the field.

Our founding principle stipulates that research and teaching are intricately entwined in our programs. We believe that the development of new ideas, insights, and findings goes hand-in-hand with their communication. SSRI emerged from two programs (MIRC and PARISS, both continuing in their affiliations with SSRI), one primarily research-driven and the other primarily teaching-focused. Similarly, SSRI has since its founding maintained an even

Duke University Photography

W

e write this introduction to the

balance between faculty and graduate students. Our strategic plan includes a major new initiative to expand this mixture to the undergraduate level. As with our graduate programs, the undergraduate programs will richly combine teaching and research. We hope to create a setting where undergraduate students will be trained in research methods and skills such that they can assist faculty in their research in a serious and meaningful way. Successful students can then go on to complete important research on their own in honors projects and independent study projects. In this way, we hope to introduce Duke’s undergraduate students to the joys of research in the social and behavioral sciences alongside faculty and graduate students.

With this edition of GIST, we announce that we are stepping down as Co-Directors of SSRI. Phil Morgan will take over as Director in the ’08 academic year. We plan a year of transitional leadership, with the three of us working together as a team in order to smooth the transition. Phil brings important skills to SSRI, given his experiences as Department Chair in Sociology, as a co-director of the second Faculty Fellows program, and as a demographer who already serves as a leader of an affiliate within the SSRI umbrella. We welcome Phil onto the SSRI staff, and we look forward to working with him to ensure SSRI’s dynamic future.

And last, while we will continue to work with SSRI into the future, this magazine provides a special opportunity to say what a pleasure it has been for the two of us to work together and what an honor it has been for us to serve the social and behavioral sciences at Duke and beyond.

John Aldrich and Wendy Wood Co-Directors, 2003-2007

gist

s p r i n g 2 0 0 8 , vo lu m e 2 , i ss u e 1

6

f ro m t h e M i l l

w w w. s s r i . d u k e . e d u / g i s t

features

10

6 The Making of an Election 10 Exploring Demographic Issues 12 Where the Environment and Health Meet

12

14 Building a Culture of Interdisciplinarity at Duke

in every issue 2 In Brief 16 Profile: Faculty 17 Profile: Student 18 Ask the Social Scientist

Editor: Courtney Packard Orning

14

Assistant Editor: Claire Cusick Designer: Regina Barnhill-Bordo www.bdesign-studio.com This publication is printed with soy inks on chlorine free paper containing 10% post-consumer fiber.

Please recycle this magazine.

19 Questions 20 Technology 21 The Strip Back Cover Upcoming Events

w w w. ss r i . d u k e . e d u 1

news

in brief

by Marci Ryan

When Fear Returns

No matter our culture, we all share some same fears. The fear of snakes and spiders, for example, has helped humans survive potentially deadly encounters. But what happens when fear—of not just dangerous things, but seemingly innocuous situations—takes over our lives? These are called phobias, and for some phobics life can be dramatically changed—and even controlled—by fear.

Cognitive therapy, a form of psychotherapy using imagery, selfinstruction, and related techniques to alter distorted attitudes and perceptions, can help people overcome debilitating anxiety and crippling fears. But for some, the effects of therapy often wane, and patients have a relapse where fears return in full force. This relapse is known as “fear renewal.”

A team of psychologists, computer scientists and neuroscientists is examining fear renewal with new investigative methods at Duke’s Immersive Virtual Environment (DiVE). Nicole Huff, a postdoctoral fellow, and Kevin LaBar, an associate professor of psychology and neuroscience, are working with Rachael Brady, an adjunct associate professor of computer science and director of Duke’s Visualization Technology Group, to unravel why fears return.

Their research could lead to identifying regions of brain activity, which could eventually target drug therapies to help phobics or patients with post-traumatic stress disorder overcome the context cues that trigger debilitating behavior.

“Fear follows you through contexts,” says Huff, who arrived at Duke in 2005 with a background in behavioral neurosciences and with a focus on animals. To study fear relapse, “First you must condition fear, then extinguish it, and test to see if it comes back,” she says. Classic laboratory environments have proved difficult to study the context of fear, so she joined LaBar, an associate professor at the Center for Cognitive Neuroscience who works with healthy humans to examine the fear conditioning paradigm, and Brady, a computer science engineer, to develop a 3D virtual reality laboratory to study fear. Their work is funded by a grant from NSF. The ability to manipulate the virtual reality environment allows the team to simulate multiple situations. They recruited healthy volunteers who have no previously reported fear of snakes or spiders. First, they simulate snakes and spiders in the subjects’ field of view. Then a group is “conditioned” to fear snakes and spiders through a series of minor shocks—only strong enough to sense, but not to hurt—when they encounter them. To gauge stress during this process, the team measures sweat production through a skin

2 gist from the mill • Spri ng 2008

Nicole Huff in the 3D laboratory

Cognitive therapy, a form of psychotherapy using imagery, self-instruction, and related techniques to alter distorted attitudes and perceptions, can help people overcome debilitating anxiety and crippling fears.

But for some, the effects of therapy often wane, and patients have a relapse where fears return in full force. This relapse is known as “fear renewal.” conductance test while subjects are exposed to fear conditioning (the method in which subjects learn to fear new stimuli) or extinction training (the study of how to extinguish or unlearn a fear).

After one day, the subjects return to the DiVE, the shock is removed, and fear responses go down. But their “safe” extinction memory is context-specific, and if presented with snakes or spiders in a following session, the fear returns. Through context manipulation—the process of changing the environment in a controlled way—the researchers are starting to understand what features drive fear.

For clinically treating fear, “timing does make a difference,” LaBar points out. “If you treat too soon it is not effective; you need time to fully form a safe memory.” Fear is also highly affected by mood. “If a person is too aroused it is difficult to unlearn fear,” says Huff. For clinical patients with PTSD, this might be especially important in how long treatment can take, or even begin, following their traumatic experience.

The next step for the team this winter is to take the virtual reality program and scan the brain with MRI while conducting the tests. By identifying the brain regions that are active during the conditioning and extinction phases, new drug therapies can be developed to treat phobics, PTSD patients and other affected populations.

Learn more about DiVE at: vis.duke.edu/dive

Duke University Photography

A Team of Researchers Examines Fear Renewal in a 3D Virtual Reality Laboratory

news

in brief

New Study Shows War on Guns is Not Lost These days, news reports can make any citizen feel unsafe stepping outside the door. Stories about shootings in a shopping mall, at a major university, or even in a church make us wonder where the guns came from, and whether they were purchased illegally.

According to a new research study published in the November issue of The Economic Journal, the war on illegal gun use is not necessarily hopeless. In fact, tightening restrictions on illegal gun sales in some cities has actually lessened gun use by criminals.

The study focused on the underground gun market in two neighborhoods in Chicago. Citywide, authorities have laid down unusually tough restrictions on handgun purchases and placed a high priority on prosecuting individuals using illegal guns. The research revealed that the difficulty involved in procuring guns has significantly impaired access to them by youths and criminals.

Philip Cook, professor at Duke University’s Terry Sanford Institute of Public Policy and co-author of the study, says one effect of the gun emphasis in law enforcement is to limit the involvement of gangs and drug dealers in selling guns. “In those neighborhoods, there are only around 1,500 illegal gun transactions, compared to millions of sales of illegal drugs in the same period,” he says.

“The question arises: why wouldn’t the drug dealers diversify their business and sell guns as well? It’s because the dealers are worried about putting their main business at risk—if they started selling guns, the police would pay a lot more attention to them.”

This forces criminals to go without a gun or turn to “brokers” acting as liaisons between sellers and buyers of illegal guns. However, this comes with a hefty payment to the broker, followed by a high markup on the firearm. Complicating matters is the fact that even professional brokers can’t complete many of the transactions they arrange.

Cook says he was most surprised to find out about the existence of these brokers, as well as the lack of knowledge criminals had about guns. “I was willing to believe there was some friction in the market,” he admits, “but I found it remarkable how difficult it was for criminals to purchase not only guns, but also ammunition. Even more so, these youths are not knowledgeable about how to get or use guns, and we would love to preserve that particular kind of ignorance.” He adds that most criminals don’t even care if the guns work or not, they just want the impression that the “bulge” creates.

One exception the study found is that gang members were better connected to

by Beve rly Sch ieman

big-time gun runners and sources outside of Chicago, and were able to obtain firearms more easily. However, the gang leaders carefully controlled the dissemination of the weapons. “They would have an arsenal where they would stash the guns, and then pass them out when needed with clear instructions on how to use them,” explains Cook. “Then they would take them back.”

Though Cook said he and his co-authors have been researching the underground gun market for years, this study provided them with direct access to information about how the transactions took place, and the difficulty criminals had obtaining guns.

“The opportunity came up when we met with Sudhir Venkatesh, an ethnographer at Columbia University,” says Cook. “For years he’s been investigating the lives of the underclass in the neighborhoods of Chicago, so we asked him to add questions to his informants about gun transactions.”

Venkatesh conducted interviews with gang members, criminals, prostitutes and the police, and actually introduced Cook to some of the people he talked to in their neighborhoods. “Sudhir is a unique individual who has the trust of the underworld figures,” Cook says.

“His approach is to find someone he can build a relationship with and meets the criminals through that person as a go-between.”

The researchers also looked at U.S. Department of Justice data from previous interviews with samples of arrestees in 22 cities.

They found that illegal gun markets in large cities with restrictions and enforcement policies like Chicago followed a similar pattern: the “thinness” of the market (minimal transactions without the benefit of advertising) combined with its illegality impaired transactions. “The ethnographic research in Chicago helps flesh out those findings,” says Cook. “These weren’t just people responding to a five-minute questionnaire in a police station; these were conversations over coffee between two people who trusted each other.”

Could this mean that a strong law-enforcement priority on getting rid of illegal guns could reduce gun use in crime? “A lot of the academic literature out there leads us to imagine sophisticated criminals and believe that guns are everywhere, but they’re not,” says Cook. “That’s the importance of our effort: it’s not hopeless or too late, and we shouldn’t give up on any semblance of control. Cities such as Chicago and New York that take guns seriously seem to be accomplishing something.”

w w w. ss r i . d u k e . e d u

3

news

in brief

by Dawn Stuart

DIISP Lab Strives to Help

Reduce Child Abuse in Durham

Research taking place at the Duke Interdisciplinary Initiative in Social Psychology (DIISP) lab is contributing to a community effort to reduce child abuse in Durham.

Lisa Berlin, research scientist with Duke’s Center for Child and Family Policy (CCFP), is conducting the study to determine how community services for at-risk families are affecting parent-child attachment. Berlin, who has been at Duke since 2000, specializes in early child development and early intervention, particularly the prediction and prevention of child abuse and neglect.

“We know a great deal about normal attachment,” she says.

“This study is bringing that knowledge together with public health and child abuse prevention.”

Berlin’s study is one of many types of social science research for which the DIISP lab was created in 2005. “We are an interdisciplinary lab,” says David Neal, lab director, explaining that the creators of the facility envisioned collaboration between branches of social science as well as achieving economies of scale.

“Having the shared facility brings people into dialogue, creating knowledge exchange and building relationships between disciplines,” says Neal.

Additionally, it prevents duplication of effort and gives researchers access to leading-edge technologies. “Using our facility and equipment can make research projects more viable and cost effective,” Neal says. “There is no need for researchers to obtain their own sophisticated audio and video recording systems.”

The DIISP lab facilities include four observation rooms equipped with state-of-the-art audio and video recording equipment, networked desktop systems and removable desk partitions for individual or group testing, and two psychophysiology recording rooms equipped with biofeedback monitoring systems.

Berlin’s technological needs for this research were fairly simple, though. She used video cameras that drop down from the walls instead of the clandestine ones. “The babies usually don’t notice them,” she says, describing the subjects of her research (children between the ages of 12 and 20 months).

Partnering with Berlin for the study is the Durham Family Initiative (DFI), a collaborative program of Duke’s CCFP and the Center for Child and Family Health, a Duke-affiliated community service provider. DFI helps Durham County families at risk of child abuse become self-sufficient and supportive of their children’s growth and development, with a goal of reducing child maltreatment by 50 percent in the next 10 years.

In the effort to meet this goal, DFI is evaluating the effectiveness of community service programs such as those of the Center for Child and Family Health. One such program, Healthy Families Durham, provides intensive support for at-risk new mothers through social workers’ home visits beginning during pregnancy and continuing

4 gist from the mill • Spri ng 2008

up to a child’s third year. Risk factors include child-bearing at a young age, mental health problems, low social support, and a history of trauma or domestic violence.

Mothers and babies involved in DFI’s evaluation of the Healthy Families program are participants in Berlin’s research involving the “strange situation,” a specific procedure widely used in studying attachment in early childhood development.

During this procedure, the mother leaves her toddler in an unfamiliar playroom for up to three minutes. Observers note and record responses to the mother’s departure, the presence of an unfamiliar person and the mother’s return.

“The situation we are creating is mildly stressful and akin to a real life circumstance such as a baby being left briefly with the nurse at a doctor’s office,” says Berlin. “The separations are three minutes at the most, less if the baby gets very upset.”

“We are interested in seeing what the stress elicits from the babies, especially in terms of using their moms as a source of comfort and support,” she says. “There is remarkable diversity in how the kids react. Not all babies are able to use their mom for support, and not all parents provide the support easily or comfortably.”

Research findings will indicate if the Healthy Families home-visit program is creating healthier mother-child attachments. “We would think that, if this program is preventing maltreatment, it should also be improving attachment,” says Berlin. “But we’re not sure. We are looking to learn more about the relationship between attachment and child abuse prevention in high-risk situations.”

Berlin has found the DIISP lab’s location at Erwin Mill among its most valuable features. “People are much more amenable to coming to central Ninth Street than a location on Duke’s campus,” she says.

“It is easy to find, and there’s plenty of parking.”

Additional resources available to lab clients are a database of 1,000 adults willing to participate in studies and $500 grants that can be used to pay participants. Any Duke department may use the facility, which is supported through the College of Arts and Sciences, the Office of the Provost and the Office of the Vice Provost for Research.

For information about conducting experimental research at the lab, contact David Neal at 919-681-6728.

news

in brief

on the job market

by Angela Spivey

Like many Duke graduate students about to receive shiny new

Duke University Photography

Ph.D.s, Jeremy Burke and Kata Mihaly are looking for faculty jobs.

In spring 2008, Burke will receive his Ph.D. in political science, and Mihaly will receive her Ph.D. in economics. The academic job market is always competitive, but Burke and Mihaly have an extra complication; they’re engaged and would like to find jobs in the same town.

That has motivated them to consider not only faculty jobs but also some economist or political scientist positions at government agencies or private companies. “We broadened our search because we wanted to increase the chances of placing together,” Mihaly says.

Another factor that will increase their chances of finding the jobs they want is that they both were fellows of the Program for Advanced Research in the Social Sciences (PARISS)—Burke in 2006-2007, and Mihaly in 2007-2008. Mihaly says that the fellowship gave her a chance to interact with researchers in other disciplines, which has made her more competitive no matter what sort of job she decides to pursue.

Though Mihaly and Burke are in different fields, they each do research that crosses the boundaries between disciplines. “Both of us approach topics that are not necessarily economic subjects but use economics tools to analyze them,” Mihaly says.

Mihaly has presented her work in the economics department, but the PARISS fellowship gave her a chance to present it to a very different group. “If people who aren’t in my field don’t understand what I say about my work, I want to fix that,” she says.

Because Mihaly studies social networks, she wanted to get the perspectives of not only economists but also sociologists, and the PARISS fellowship was perfect for that. “The fact that I study individuals who are deciding to make friendships and deciding how much effort to put into school—that gets into individual utility maximizing, which is an economic concept,” she says. “Networks have been widely studied under sociology but are only recently being studied in the economics field.”

Burke uses game theory, which is typically thought of as an economics tool, to analyze the choices that media outlets make.

“Game theory started out as mathematics, but has been widely used in economics,” Burke says.

Basically, he asks the question, how does competition affect the quality of information that people receive Jeremy Burke and Kata Mihaly from the media? “You can think about it in terms of, there are two rules for media firms.

One is to be an arbiter for truth and tell the population what’s happening. But of course they also have to maximize profits to satisfy shareholders,” Burke says.

Data Services Core Brings Researchers Together Venture down a small hallway in SSRI’s space in Erwin Mill

Duke University Photography

Building, and you’ll come across a handful of rooms (and employees) housing a very big idea. The Data Services Core (DSC) is a new Duke University social sciences initiative to bring researchers together across disciplines, and help them meet their goals.

The DSC assists faculty and students with data tasks that might be daunting to tackle on their own, including acquiring and managing access to large data sets for secondary analyses, designing large-scale surveys, analyzing statistical data obtained from research, and organizing that data into charts, maps, and other graphic

by Beve rly Sch ieman

forms that can be published or presented.

“It’s a centralized resource focused on data sets that might be substantially larger than what individual investigators might obtain on their own,” says Rick Hoyle, Research Professor of Psychology and Neuroscience and Director of the DSC, “particularly data that might inspire collaboration, because our mission is to bring multidisciplinary groups of researchers together and provide the environment and support for them to do cutting-edge research.”

The DSC also provides workshops for social scientists and students on topics such as statistical methods, data-analysis software and working with databases. According to DSC Associate Director Gary Thompson, the DSC operates along all points on the research spectrum. “We can help them figure out what kind of data they need to obtain, track it down, and give it to them in a statistically relevant way that would be useful,” he explains. “Often there’s |data out there that’s not so easy to find, and faculty researchers are busy people. They could find it, but it’s faster for us to find it.”

For more information, see: ssri.duke.edu/dsc

Gary Thompson and Professor Rick Hoyle w w w. ss r i . d u k e . e d u 5

6 gist from the mill • Spri ng 2008 8

The Making of an

Election

by Nancy E. Oates

The presidential election process has changed since the men attending the Constitutional Conference in Philadelphia decided George Washington would be the best man to serve as the first president of the United States.

Because no one opposed his selection, he did not have to run a race, obtain party backing, stump for the support of the in 2008 marks the first time in more than 50 years that both major parties have contested elections. Here, three Duke researchers examine the complexity in the electoral process that has evolved over the past 219 years.

The Iowa C aucuses

Iowa residents overwhelmed their precinct sites the night of the Jan. 3 caucus, turning out in more than twice the numbers expected. Parking was scarce; overworked registrars worked steadily, but even so, when the doors closed at 7 p.m. to begin the caucus, those still standing in line were shut out.

John Aldrich, Pfizer-Pratt University Professor of Political Science at Duke, observed a high school auditorium in Iowa City. He was impressed by how engaged caucus-goers were in electing non-binding delegates to a convention that would

not be held until after the candidates had been virtually selected by primaries in the remaining 49 states.

“I could trust my democracy to these folks,” Aldrich says.

Aldrich made the trip to observe the caucus process and to present his ideas on campaign reform at a conference at the University of Iowa the following day. Since 1972, the Iowa Caucus has been a predictor of which candidates will ultimately represent their party in the presidential election in November. Caucus-goers can support their second-choice candidate if their favorite can-

l to r: Duke Professors John Aldrich, Mike Munger and Alexandra Cooper, Ph.D., are studying different aspects of campaign finance during the 2008 elections.

didate doesn’t make it among the top three.

Candidates who receive negligible support in Iowa often drop out of the race, reducing the choices in the primaries to come.

But this year, the Iowa Caucus fell victim to the rush of states to be first. The first few states to hold electoral events benefit from increased exposure and tourist dollars from candidates, their entourages and the media. Traditionally held toward the end of January, the Iowans rescheduled to Jan. 3 to stay ahead of New Hampshire’s primary, which moved itself up to Jan. 8, from its endof-January date in 2004. Several states folw w w. ss r i . d u k e . e d u 7

Duke University Photography

general populace or raise money. The presidential election

lowed suit: 24 states hold primaries on Feb. 5, the first day the Democrat and Republican parties allow delegate selection. (The parties have granted exceptions to Iowa, New Hampshire, South Carolina and Nevada.)

In the paper he presented at the campaign reform conference, Aldrich argued against such front-loading. He proposed that the window for selecting delegates be open only in April and May to avoid a six-month dearth of political activity until the national conventions in August. In 1992, Ross Perot took advantage of that hole to launch his campaign as an independent. By the time the national conventions rolled around, he was in a three-way tie with Bill Clinton and George Bush.

Aldrich also proposed augmenting the amount that candidates could receive in federal matching funds. The heft of candidates’ campaign coffers has been used as a bellwether for success. Candidates who accept federal matching funds of up to about $21 million must agree to a campaign spending limit of $50 million. Sens. Hillary Clinton and Barack Obama passed up matching funds.

Each raised more than $100 million in 2007.

“I proposed matching grants be raised up to 110 percent of what was spent in the last presidential campaign,” Aldrich says. “This would make it possible for lesser candidates to raise enough money to be competitive.”

8 gist from the mill • Spri ng 2008

He also urged that those funds be spent only during the two-month window for delegate selection, plus one month before the window opened. That would effectively limit spending to the campaign itself, rather than allow potential candidates to spend money raising their profile during the year leading up to the primaries.

“That invisible primary is of equal value to winning Iowa and New Hampshire together,” he says.

Donor Influence

Presidential campaigns consider small donors to be different from big donors.

They have less money and a few more years on them, according to a comparison of data collected in 2000 and 2004 by a trio of researchers through SSRI, but are as disproportionately white, male, old, educated and wealthy as the big donors. In between the two surveys, in 2002, the McCain-Feingold Bipartisan Campaign Reform Act passed, doubling the limit individuals could give to $2,000, leading to concerns that even more of the money presidential candidates raise comes from an unrepresentative sample.

Researchers noticed that small donors, contributing less than $200, seemed to matter more in the 2004 election, said Mike Munger, chairman of Duke’s political science department.

“It’s as if there’s this whole continent where there are these creatures that no one has ever seen before,” said Mike Munger.

Munger, along with John Aldrich, and SSRI’s associate director for education and training, Alexandra Cooper, created the campaign donor survey sent out to a random sample of 6,000 donors in 2000 and again in 2004. The researchers are preparing to send the survey out again after the 2008 election, the first presidential election since 1952 that is truly contested for both political parties.

The survey asks donors about their political views, whom they gave to, who asked them for money, and how they decided who to give money to, and the many factors involved in the decision, such as whether the candidate is friendly to the donor’s profession or industry.

“We’re all concerned whether money has an impact on elections and political decisions, but it is a distasteful topic,” Cooper says. “But it’s an important topic that matters to us if we care about how our democracy works.”

Individual donors contribute to the success of presidential campaigns more than organized interest groups do. Understanding which individuals give money and how they choose candidates to support could shed light on the choices the public will have when elections roll around, Cooper says.

By collecting the data, the researchers are building a body of information that could be useful to policymakers thinking about regulation, activists thinking about how money skews political results, and fundraisers thinking about effective strategies, Cooper says.

The researchers aren’t looking for any particular trends or implications, Munger says.

The survey simply documents behavior.

“Data collection is more important than you think,” he says. “More people are interested in it than you can ever imagine.”

Ballot Access

For Mike Munger, in his role as Libertarian candidate for governor of North Carolina, victory would be 2 percent of the popular vote. But first, he has to get on the ballot. A law passed by the state in 1982 aims to keep him, and all other independent candidates, off the ballot.

North Carolina is second only to Oklahoma in restricting ballot access to third-party candidates. The 1982 law requires candidates to collect about 100,000 signatures, which will be winnowed down to the 67,000 verified valid signatures needed to get on the ballot, and receive 10 percent of the popular vote for their party to stay on the ballot in subsequent elections. House Bill 88, passed in August

2006, reduced the latter component to 2 percent.

“I can get 2 percent,” Munger says. “I can get the Libertarians on the ballot. It’s still wrong that ballot access is so restrictive.”

An expert on ballot access law, Munger testified in the U.S. Senate in 2000 against the McCain-Feingold Bipartisan Campaign Reform Act and in November was an expert witness in a case brought by the American Civil Liberties Union against the state claiming the ballot access restrictions violated North Carolina’s constitution. The attorney general contended that the restrictions were necessary to reduce “clutter” on the ballot that might impair voter’s ability to choose.

Munger rebuts the clutter argument by citing the 2003 gubernatorial election in California, won by Arnold Schwarzenegger.

“He was one of 153 candidates,” Munger says. “People were worried that it would be an illegitimate election because the winner could win with 15 or 20 percent of the vote, because the other votes would be scattered among so many people. Schwarzenegger got 48 percent of the vote; the guy who came in second got 30 percent.”

Research, some of it conducted at Duke, shows that in states with easy ballot access, Democrats and Republicans are much more responsive to voters and there is less

Michael Tofias

“ We’re all concerned whether money has an impact on elections and political decisions...it’s an important topic that matters to us if we care about how our democracy works.” —ale xandra coope r

corruption. But the majority parties don’t want that competition.

Munger compares it to Coke and Pepsi being asked if other soft drinks should be allowed to compete with them. Not surprisingly, Coke and Pepsi would say no, he says.

Continuing the analogy, he posits asking people if they’d sign a petition to get more choice in soft drinks. But they couldn’t taste the new drink first, because he can’t sell it until he gets 100,000 signatures. “It’s very hard to get people to sign these petitions, if they have no idea what it’s about,” he says.

The Libertarian Party, which has appeared on the ballot in eight previous elections, will get its 100,000 signatures. “We will have spent all our money and arrive breathless at the starting line, having no real chance to run an election, because we will have dissipated all our resources getting the signatures,” Munger says. “That’s what the law is designed for.”

Third parties might well draw votes from the Democrats and Republicans. That would indicate voters were expressing their political dissatisfaction, a right the First Amendment grants.

“The people, not the government, ought to decide what positions are represented in the political spectrum of parties,” Munger says.

“That’s democracy.”

w w w. ss r i . d u k e . e d u 9

Exploring Demographic Issues The multidisciplinary field of

demographics lies between the social and biological sciences, and rests on a bedrock of mathematics and statistics. When policymakers need solid information about issues such as population, fertility and aging to inform their decision making, they look to demographers. And often, they turn to work from the Duke Population Research Institute (DuPRI), an interdisciplinary research organization that is an SSRI affiliate.

At DuPRI, researchers from a range of Duke departments—including the biological, mathematical, statistical, social and policy sciences—come together to explore demographic issues. While the goal of each of the five centers housed at DuPRI is to conduct high-quality research that contributes important findings to the field of demography, each center focuses on a specific niche:

10 gist from the mill • Spri ng 2008

The Population, Policy and Aging Research Center (PPARC) supports a wide range of research on aspects of aging, morbidity, mortality and longevity. The center’s overarching goal is to explore the nature of, and limits to, the plasticity of longevity, or the differences in aging according to varying environmental conditions.

The Center for Research on the Evolutionary Demography of Aging (CREDA) promotes a synthesis between the fields of demography as practiced by ecologists and evolutionary biologists and as practiced by human demographers.

Chinese Populations and Socioeconomic Studies Center (CPSES) promotes national and international scholarship on studies of contemporary Chinese populations and societies, and promotes training in the fields of Chinese population and socioeconomic

by Michele Lynn

studies. One of its goals is to establish a research and training network for scholars interested in Chinese population and socioeconomic studies around the world, in order to enhance scientific exchanges and collaborations.

Center for Social Demography and Ethnography (CSDE) examines how culture and ethnic identity and affiliation affect various family formation patterns, fertility and related matters.

Center for Population Health and Aging (CPHA) focuses on the intersection of longevity, health, disability and aging, paying particular attention to factors from childhood into old age that affect longevity and healthy aging. The goal is to study factors that contribute to longer life spans with lower levels of disability and better overall health.

“Duke is one of the few programs that has demographers associated with its public policy program.” l to r: Professors Ken Land, Linda Burton, Phil Morgan and Jim Vaupel

To learn more about the work of DuPRI, we spoke with Jim Vaupel, chairman of the board of DuPRI and faculty director of the Population, Policy and Aging Research Center, and Ken Land, faculty director of the Center for Population Health and Aging.

“The research that we have done at Duke has had a major impact on people’s understanding of how long people are going to live and the increase in life expectancy over the past 100 years,” says Vaupel. “Our work has shed a lot of light on population aging; the new world that will be created as larger and larger numbers of people are older; and how this will change society and impact public policy. We contribute truth and research findings that help inform the public policy process.”

While Duke has great strength in mathematical demography and sociology and demography—a traditional area of demographic research—the university has more recently developed strength in economics and demog-

raphy. Duke is one of the few schools that has demographers associated with the public policy program. This is allowing DuPRI to reach out in new directions to public policymakers.

Duke is also branching out by bringing biologists together with demographers to examine questions such as the male/female health survival paradox (men are healthier than women according to certain objective factors, yet men die first). DuPRI researchers are working to understand to what extent this paradox is due to social and behavior factors vs. biological factors.

“Another major thrust is working with global health initiatives,” says Land. “A lot of the work we do in demography is healthoriented. Once one begins to speak about improving global health, one necessarily needs to speak to demographers who can measure the population’s health, both globally and locally.”

By its very nature, demography is an ex-

tremely interdisciplinary subject that touches on all aspects of the social science disciplines. And while much of the work is based on the analysis of large quantities of data, there are many demographers focusing on qualitative aspects of research.

Vaupel says that the policy implications of DuPRI’s work are important as researchers provide policymakers with information about what is happening and what is likely to happen, at least in the near future. He adds that the research also has an impact on individual lives. “Most young children today will live into their 90s and may celebrate their 100th birthday,” he says. “Our research gives people facts about what their life course will be like, and they can use that information to plan their lives. Research showing that people are having fewer children and that children are leaving home later also gives individuals information that they can use in their planning.”

“Our work has shed a lot of light on population aging; the new world that will be created as larger and larger numbers of people are older; and how this will change society and impact public policy. We contribute truth and research findings that help inform the public policy process.”

—Jim Vaupel w w w. ss r i . d u k e . e d u 1 1

Where the

Environment and Health MEET

by Karl Leif Bates

When it Comes to Human Health, the Smoking Guns Aren’t Always Physical

To better understand human health risks and the factors that cause disease, Duke researchers are joining forces to find the places where obesity, diabetes and asthma meet real-world environmental conditions like rodent droppings, lead paint and social stress.

They want to know specifically where and how we get sick in the hope of being better able to prevent us from getting sick in the first place.

“Health is not just about pathogens,” says Philip Morgan, Chair of the Duke Department of Sociology. “There’s also a sense of wellbeing: Stressors matter.”

Duke researchers from sociology, the Nicholas School of Environment and Earth Sciences, the Medical School, and the Center for Child and Family Policy are combining the latest tools of environmental science, such as geographic information systems (GIS),

12 gist from the mill • Spri ng 2008

with social and health statistics to create rich new understanding.

“Duke is superbly set up to take advantage of this,” says Nicholas professor Richard Di Giulio. “We need to take a step back and look at how the larger environment affects human health,” says Di Giulio, who directs Duke’s Integrated Toxicology Program and Superfund Basic Research Center.

At the core of much of this activity is the Children’s Environmental Health Initiative (CEHI), a family of ambitious research projects looking at how, why and where children’s health varies in relation to such

things as lead exposure, indoor air quality, and prenatal care and conditions.

The growing frontier of this work centers on the neighborhood environment—sources of stress or unhappiness, for example, that are difficult to measure but that have real health consequences. Numerous studies are finding distinct correlations between negative experiences, like being on the receiving end of discrimination, and poorer health indicators.

Living in an unsafe community is itself a kind of toxin, argues Duke psychologist Kenneth Dodge, who directs the Center for Child and Family Policy. Verbal and physical abuse are both what he calls “environmental toxins,” with a growing body of epidemiology to back him up. “These are psychosocial toxins.”

If you add to this map of psycho-social stressors the sorts of factors that poor housing contains—lead pipes, lead paint,

The growing frontier of this work centers on the neighborhood environment—sources of stress or unhappiness... that are difficult to measure but that have real health consequences. Numerous studies are finding distinct correlations between negative experiences, like being on the receiving end of discrimination, and poorer health indicators.

poor air quality—there’s a remarkable overlap, Dodge says. “It turns out those things are correlated, and these are often the very same families.”

Dodge is currently working with CEHI Director Marie Lynn Miranda, an associate research professor at the Nicholas School, on the geography of child abuse in Durham. Using a Durham County database of reported abuse cases, they are building a map of where children face the greatest dangers. If the abuse pattern matches with other mapped factors, like lead pipes and substandard housing “is it something about these people that are attracted to this neighborhood, or is it something about the neighborhood that does this to people?” Dodge wonders.

That’s part of why Duke cardiologist Robert Califf has been speaking with Miranda about a collaboration. To him, Durham is a very promising laboratory for finally getting at the complex and subtle environmental factors that make people ill.

Califf, the vice chancellor for clinical research at Duke Translational Medicine Institute has made a career of sifting through huge data sets from clinical trials and hospital care to look for subtle, but telling patterns in health outcomes. He thinks that if it could be turned into a giant, multi-variable database itself, Durham represents a reasonably representative sample of the United States as a whole. It helps his plans, Califf adds, that Duke Translational Medicine Institute has taken on the responsibility for both major hospitals and the majority of outpatient care in Durham County.

What he’s begun advocating for is the establishment of fully electronic patient records for Durham residents that could then be sliced in a thousand directions for research analysis—under very rigorous

privacy protections—to reveal patterns of environmental health effects.

Marie Lynn Miranda also is known for her big picture work. Really big picture.

“I conceive of the data visually,” Miranda says, explaining a colorful poster outside her office that depicts dozens of environmental variables overlaid on a map of the city of Durham.

Technology helps researchers

Just down the hall, it’s North Carolina hot in the room where Alicia Overstreet works as a data analyst for the CEHI project. The air conditioning is actually working fine in the Levine Science Research Center on this morning; but this room contains enough computing power to plan a moon mission.

For security reasons, CEHI’s intensive computer environment is cut off from the rest of Duke’s data infrastructure, and it draws enough electricity on its own that it once started a small fire and blacked out most of the LSRC. Powerful computer work stations with two large flat-panel monitors each are used to crunch all the data coming in from CEHI’s many projects and map the information into an increasingly rich, house-by-house tapestry of Durham and county-by-county analyses of North Carolina’s 100 counties.

In the next room, there’s a printer than can produce a 3x5-foot poster in thousands of colors.

As difficult as it can be to obtain good population-level health data to build these maps, that effort pales in comparison to collecting the fine-grain data these environment-meets-bloodstream studies demand.

The rubber hits the road in the CEHI datagathering effort called the “built environments team.” Each member of the threeperson team carries a GPS data logger,

essentially a king-sized, bright yellow Palm Pilot that tracks its location within inches while team members slowly move up one street and down the next, carefully documenting conditions at every one of Central Durham’s single-family homes.

“We’ve seen bullet casings before,” says Duke senior Tomas Will.

“We’re trying to get at the housing quality, but also the things that indicate a sense of ownership, a sense of agency,” says Yates Coley, a 2006 Nicholas grad. “Missing screens, ‘no trespassing’ signs, what do they mean?”

The team hits the streets early each morning to beat the sun. They are covering every parcel in 22 Durham neighborhoods, standing at the curb in front of each house for less than a minute and recording dozens of data points on each with rapid tapping of a plastic stylus on the data logger’s screen.

Broken windows? Click click. Peeling paint?

Click. Sidewalk in good shape? Is there a porch? Click, click-click, and on to the next house. Each afternoon, they download the day’s clicks into the burgeoning database.

In a good week, a four-person team can cover 1,000 homes.

“I can see how we’ll connect this to childhood health,” Coley says, standing next to two tires and an automobile gas tank that have been piled next to the curb for months, by the look of them. Next door, a three-wheeled car leans on a rickety jack with no one in sight.

Everyone involved in this work holds out the hope that it will lead to better understanding, and truly effective prevention, for health disparities among Americans and some of their most vexing health problems.

The solutions may well lie more in the environment than the health care setting.

w w w. ss r i . d u k e . e d u 1 3

Building a culture of interdisciplinarity at Duke

Q&A with Peter Lange

by Sylvia Pfeiffenbe rge r

SP: Why did Duke make interdisciplinarity a priority?

PL: There were multiple reasons. The first was recognition that there’s a lot of intellectual excitement at the boundaries between disciplines, and that if you could create a culture in which those boundaries would be explored, you could achieve intellectual leadership in some areas. Two, Duke is a middle-sized research university with absolutely top-notch aspirations. By exploiting interdisciplinarity, we can build communities that are larger than departments by bridging across. Third, we’re an adventurous place. We’re a place that takes a lot of pleasure in innovating and trying new things. I think that’s part of the culture and character of the campus.

SP: What has Duke done to merit national recognition in this area?

PL: The real cutting edge is whether you can drive your commitment to interdisciplinarity into the actual structures by which decisions are made. We’ve been very, very persistent in pushing our agenda, putting the resources behind it, identifying the obstacles and removing them one after the other. That doesn’t mean we’ve removed them all, we have a bunch to go, but we’re penetrating deeper and deeper into the fabric of the university. And we’ve had considerable success in that we’ve been able to recruit outstanding faculty and attract really good students, and do really good research around this commitment to work across disciplines.

SP: What kind of obstacles have you identified?

PL: Some of them are in our minds, but many of them are bureaucratic, they are in rules, they are in budget, they are in allocations, they are in time. You have to just keep eroding the obstacles, identify-

14 gist from the mill • Spri ng 2008

ing what they are, seeing if you can change them, or if you can’t [change them], work around them, and also creating the visibility and the commitment. Duke was one of the first places to create a Vice Provost for Interdisciplinary Affairs, and that is a very significant, visible commitment. The fact that we made interdisciplinarity such a central focus of our strategic plan, and that we put money behind commitments, is another one of those.

SP: What kinds of bureaucratic obstacles have you been successful in removing?

PL: It used to be the case that when collaborative research was done, that the overheads from federal grants went to the school of the principle investigator, so if he or she had collaborators in other schools, too bad.

We changed the rule, so now the overheads break into two pieces, one facilities piece and one administrative piece. That removed an obstacle. It’s where is the work being done and by whom, no question about what school the P.I. is in. People can collaborate and not worry about the allocation because we created a system of just distribution.

Another example is that there were difficulties about professorships in multiple schools or even in multiple departments.

We changed the rules to make that easier.

There’s a memorandum of understanding that’s written between the two departments, or centers, or institutes, about what division of responsibilities and labors the person will have. Now an enormous number of our faculty have secondary appointments, and are doing work in other departments.

SP: How do you bring about changes in people’s attitudes?

PL: What happens is this: well you know, I thought it was a pain in the rear to do this,

but now I see how much it benefits us. And oh look, our reputation is bigger, or, I have a much better research community, or look at all the colleagues I have, or look at the things I thought about doing but I couldn’t do at the school I used to be at that I can do here. So that sets up a virtuous circle, it begins to change the culture.

SP: How much of the success depends on the design of physical spaces, i.e. having office spaces nearby, public areas to interact in, and so on?

PL: I think the configuration of physical space is significant. In our new buildings we’ve paid a lot of attention to that, because there are public spaces there, spaces to interact and be comfortable. In an older building, if everybody is behind brick walls, literally, you have to invade somebody’s territory to find out if they’re there. If you have a glass wall, all of a sudden you don’t have to invade. And it makes a big difference. We’re sort of learning as we go along.

When we touch old buildings, we are trying to bring these new ideas into being, and in the new buildings we’re obviously trying to build them in.

SP: What has been the impact on curriculum? How does this go beyond just cross-listed courses?

PL: Now we have lots of certificate programs which are interdisciplinary. Not every course has to be interdepartmental. Let’s say it’s neuroeconomics. You might take courses in economics, psychology, and then you might also take a course in brain science. But there’s a capstone course that is going to bring those things together, and an introductory course that is going to show you that intellectual world where the disciplines come together.

Duke University Photography

“ Interdisciplinarity is a craft, not an art. You don’t make it up originally every time, but it’s also not an industrial product, you can’t just reproduce it.”

—pete r lange

Provost and Professor of Political Science, Peter Lange leads a discussion on interdisciplinarity at Duke.

SP: What else has Duke learned about doing interdisciplinarity right?

PL: Well, we’ve learned it’s hard. The metaphor I usually use is that interdisciplinarity is a craft, not an art. You don’t make it up originally every time, but it’s also not an industrial product, you can’t just reproduce it. Every interdisciplinary unit has its own particular features; everyone has to be crafted in a slightly different way, because you’re building communities among people that are not forced to be together. Different communities have different requirements and you have to be very sensitive to those.

SP: What’s the payoff?

PL: It’s fun. You’re putting energy into your system. You have people on the ground saying that they feel empowered. It’s a carrot for both recruitment and retention.

We’ve been able to recruit fabulous faculty members from very strong places because they could do collaborative work that they couldn’t do at the school they were at before. Once you’re here and you build those interdisciplinary communities and you know you can’t get them elsewhere, it’s more of an inducement to stay. You don’t only attract faculty with money. You have to pay them well, but on top of that, it’s the ties that bind.

the strategic plan, which led to the creation of these. So now [that] we’ve created them, [we need to focus on] what are the next steps to making them really integral.

Last year and this year together we’ve been creating the capability so that those units could hire regular rank but non-tenure-track faculty such as research professors. You have to create a set of rules to assure that those people are well-protected in those positions.

At the same time you have to be sure you don’t create rigidities. So we have to work all that out. The IGSP applied for that last year.

Now we’re moving ahead and a couple of more will be applying for that authority over the next three or four months.

Also, the Sanford Institute is on a path to become a school in the next couple of years. Upgrading that unit is another commitment, because public policy itself is an interdisciplinary field.

SP: Where are we going, and how will we know when we get there?

PL: There’s no end point. It’s always a balance thing. We’re not abandoning disciplines. We’re continuing to support departments, tenure and tenure-track appointments are still made in departments and not in the interdisciplinary institutes or centers.

One of the purposes of the latter, is to insert flexibility into the system and if you start hardening them up, you end up replacing the old system just on a different frame.

The fact that it’s taken us 20 years doesn’t mean that we’ve made all the progress we want, but it also means we’ve made a lot. We already have years of work we’ve been doing to build a culture. We got lucky, or we were smart, one or the other, or more probably both, to be in the forefront.

SP: What goals are on the horizon?

PL: What we do every year is target a few of the structures that obstruct progress and that we need to shift, soften, or eliminate.

This year we’re working very hard on these seven signature interdisciplinary institutes and centers. Susan Roth, Vice Provost for Interdisciplinary Affairs, is leading the charge on that, and devising administrative and budgetary models and oversight models that will enable them to be as dynamic as they can be. Last year we had

w w w. ss r i . d u k e . e d u

15

facult y

profile

History through Baseball

by Mary-Russell Robe rson

“Whoever wants to know the heart and mind of America had better learn baseball.” That was the pronouncement of Columbia history professor Jacques Barzun several decades ago.

Today, says Duke history professor John Thompson, “Baseball is no longer America’s sport. I would say whoever wants to know the heart and mind of America should hang out in a bar on Sundays and watch professional football.”

Still, baseball was king for more than a century. And according to Thompson, we have much to learn from the sport. In fact, he’s teaching a history course this spring called Baseball in Global Perspective. “It’s a way to teach international history and connections between the United States and the world,” Thompson says.

“Baseball has become a world sport.”

Baseball had its origins in an English children’s game that came to North America with migrants in the 1700s. It the early 1800s, it became popular among young men in colleges and the military in the northeast and in Canada. The rules were still evolving and varied from league to league: some teams used five bases instead of four, and others counted a ball caught on one bounce as an out. Between 1857 and 1864, the National Association of Baseball Players standardized the rules, and modern baseball was born.

Soon thereafter, baseball came to be thought of as America’s sport. “The other country that invented baseball was Canada, but in the late 19th century, baseball became the quintessential

Professor John Thompson says baseball is a mirror for America’s history and culture.

Duke University Photography

16 gist from the mill • Spri ng 2008

Americanism,” Thompson says. “The people who sold sporting goods and owned baseball teams worked hard to associate baseball with Americanism.” Baseball later spread throughout the world on the heels of economic contact.

Thompson is a Canadian himself, having grown up on the prairies outside Winnipeg, Manitoba. He played baseball as a boy, and later picked up the saxophone. In high school, he performed with a couple of different bands and was acquainted with another young man his age from Winnipeg who fronted a band called Neil Young and the Squires.

While Young followed his calling to become a rock ‘n’ roll star, Thompson followed his calling to become a historian. His office on the third floor of Carr is crammed with books, framed photographs, and memorabilia relating to history, baseball, and Canada—and he’s happy to talk at length about any combination of these.

In addition to teaching the baseball class, Thompson is working on a book about North Carolinian baseball player Enos “Country” Slaughter. Slaughter was born in Person County in 1916 and played for the St. Louis Cardinals for years. He coached Duke’s baseball team from 1971-1977, and died in 2002. “He is an interesting lens for looking at celebrity, masculinity, gender and race,” Thompson says.

In 1947, the Brooklyn Dodgers signed Jackie Robinson, the first African-American to play in the major leagues. The Cardinals, the story goes, planned to strike in protest. The strike never happened, but when the Cardinals played the Dodgers, Slaughter spiked Robinson on the leg with his cleat as he ran past first base.

That story may be common knowledge, but Thompson is digging for facts. Northern newspapers published rumors of a boycott, but Thompson says, “I have not found a shred of evidence that the Cardinals considered striking.” Was Slaughter’s spike a statement? It’s hard to say, as Slaughter spiked opposing players frequently. (“The circumstantial evidence indicates that a better nickname for Enos than ‘Country’ would have been ‘Spike,’” Thompson says.)

“Enos became the bad guy to Jackie Robinson’s heroic good guy,” Thompson says. “You can’t have a good guy if you don’t have bad guys. He spent the rest of his life trying to demonstrate he was not racist.”

To some, baseball is just a game. To Thompson, it’s a sport and a business that mirrors America’s history and culture. “Baseball can tell you about race,” he says. “Baseball can tell you about gender.

Baseball can tell you about class. Baseball can tell you about the history of business in America.”

student

profile

Efrén Pérez Student feels welcome at Duke, but probes whether Latinos are welcome in the U.S.

by Clai re Cusick

“It is often said that all research is biographical. My case is

Duke University Photography

no different. Each of my projects is informed and shaped by my upbringing in an immigrant home.”

Duke doctoral candidate Efrén Pérez, the son of Mexican immigrants, wrote that passage on his personal web site, and indeed it’s true. His dissertation “investigates the extent to which Americans are anti-Latino rather than anti-immigrant. Using survey-experiments and time series analysis, I assess the degree to which racial misgivings about Latinos are expressed through support for stricter immigration policies.”

Professor Paula McClain, who is Pérez’s adviser, says Pérez’s research answers an important and complex question. “The various methodologies he is using, e.g., experiments and time series analysis of survey data, will give him more stringent tests of his hypotheses than any one technique alone.”

“His work is very exciting and will have an effect on the broader discipline of political science in general and race and politics in particular,” she adds.

Pérez, who will receive his Ph.D. in May, was always intrigued by politics, and became interested in political science as a possible career during his undergraduate years at the University of San Diego.

“I needed to figure out what I was going to do,” he says. “My mom wanted me to be a pharmacist.” He took an international relations course with a professor who he describes as very animated. He asked the professor how one went about becoming one, and the answer was, “You’re basically a student for the rest of your life.”

Well, Pérez thought, I’m a good student. “If it ain’t broke, why fix it.” He decided to pursue a Ph.D., but not immediately after earning his bachelor’s degree. He wanted some real-world experience first.

So he headed east to Washington, D.C., where he served as a Public Policy Fellow for the Congressional Hispanic Caucus Institute. Then, he returned to his native Los Angeles, and worked as a media relations advisor for a political consulting firm. He valued both jobs. “[Media relations] was good in many ways; it let me know what else I could do in case graduate work didn’t work out,” he says. It also gave him good practical skills, he says. Learning to write under pressure helped him later, when he had to set his own pace in graduate school. “I set deadlines for [myself ], and worked incrementally, being my own taskmaster.”

Even with that experience, and the long-held intention to earn a Ph.D., Pérez says he didn’t know exactly what he would study upon arriving at Duke. “I think any person going into graduate school, unless you have uncommon foresight, you come in kind of naïve about what you’re going to do,” he says. “It’s part of the process.” The reason he came to Duke was the Political Science Department’s collegial atmosphere, which he could sense while still an applicant.

During his time at Duke, Ph.D. candidate Efrén Pérez says presenting his work at a REGSS colloquium helped him focus.

“The folks responded; faculty and students alike,” he says.

“I took that as a sign that the culture would be different here, less hierarchical, more collegiate. I guess I decided well in that regard.”

During his time at Duke, Pérez has worked with the Center for the Study of Race, Ethnicity and Gender in the Social Sciences (REGSS), an SSRI affiliate. This experience bolstered both research and collegiality, he says. “I’ve had an extra plus. I’ve had two intellectual homes, political science and REGSS,” he says. “What looks one way in political science might look differently in REGSS. What REGSS gave me was the early work that eventually turned into my dissertation.”

He first presented that early work at a REGSS colloquium. He got a lot of usable feedback, and areas to work on. It helped him focus, he says. “Rather than stalling a year, I was able to make great strides in my research,” he says.

Pérez credits his parents’ work ethic for inspiring him. His father worked in a kitchen for 30 years, and his mother worked various office jobs and cleaned houses on the side. “Dad went to third grade, Mom went to sixth,” he says. “Their expectation for all their kids was, ‘You’re going to go to school. You need to go to college. You need to do better than we did.’” His and his siblings’ success in life means “their efforts weren’t in vain.”

Pérez also credits his advisers for his success. “If I’ve been quote-unquote successful, it’s because people have shown up in my life,” he says. “You need active and supportive environments.

You need advisers who care.”

“The center, in its own way, gave me the confidence to share my work,” Pérez says. “Research is very private and even very biographical … It’s your baby in some ways. But [sharing and receiving feedback] is definitely the benefit of REGSS. It gives you that forum.”

w w w. ss r i . d u k e . e d u

17

a sk th e social sci e ntist

Questions Submitted by Duke Undergraduates The Rolling Stones have not had a hit in decades, J.D. Salinger has not published anything since 1965, and Orson Welles was just 26 when he wrote, directed, and starred in Citizen Kane.

Why does it seem that people reach their creative potential so early in life?

Research shows that, in most fields, people make their greatest professional contributions while in their 30s and 40s. Professional productivity differs somewhat by field, but in general, 40 to 50 percent of people’s total professional contributions occur in their 30s and 40s, and they tend to make their most influential career contributions when they are in their 40s. After that, the productivity of people in the arts drops off rather sharply, whereas scientists and other scholars show relatively little decline in productivity until their late 60s. So, for example, artists make only 5 percent of their total contributions in their 70s, whereas scientists make 15 percent of their contributions after 70. Even though Mick Jagger is still performing at 64, the fact that the Stones made their major contributions to popular music when the band was younger fits the general pattern.

Why do so many powerful political and business leaders, including not a small number of U.S. presidents, seem to have so many extramarital affairs?

Of course, data do not exist to document the frequency of affairs among male business and political leaders and whether it differs from less powerful men. However, if your assumption is true, the answer may lie in one known correlate of power motivation in men.

Overall, the psychological and behavioral correlates of having a high desire for power are mostly the same for men and women.

However, one difference involves the fact that, among men (but not among women), having a high motive for power and influence is associated with a pattern of reckless behavior that includes profligate, impulsive actions. Research has suggested that this effect is due, in part, to differences in how men and women are socialized with respect to responsibility and caring for other people. Highly powerful men who are highly responsible do not show the same level of impulsive sexuality as those who are less responsible.

18 gist from the mill • Spri ng 2008

Duke University Photography

Mark Leary

Why do people blush?

Two primary explanations of blushing have been offered. One suggests that blushing works as an involuntary nonverbal indication that a person recognizes that he or she has violated some norm or standard and, thus, feels badly about it. In some ways, it functions like a nonverbal apology. The other view suggests that blushing is a response to undesired social attention and helps to divert others’ attention away from the blushing person. This explanation offers more generality in that it explains why people blush even when they have not violated any rules, such as when they are singled out for excessive positive attention – for example, having “Happy Birthday” sung to them, or being stared at by others. The nonverbal behaviors that typically accompany blushing, such as gaze aversion and a silly, mirthless grin, resemble the appeasement displays seen in other primates. Like primate appeasement, blushing and its accompanying expressions reduce negative reactions by other individuals.

When people form close relationships, they seem to seek people who are similar to them in some ways but quite different from them in others. What does research show regarding the processes involved in relationship development?

On most characteristics, people form their closest relationships with those who are more similar than different. On the vast majority of characteristics that have been studied—demographic variables, personality traits, attitudes, and physical attractiveness— birds of a feather tend to flock together. For example, engaged and married couples are significantly more similar on most characteristics than random pairs of people. On only a small handful of attributes do people sometimes prefer those whose characteristics are notably different from their own. In these instances, it’s not a matter that “opposites attract”—for example, nice people don’t prefer mean people, and clean and messy people typically don’t get along. Rather, people are attracted to those who are different from them when the difference involves attributes that are complementary. For example, highly dominant people may prefer somewhat more submissive friends and partners, whereas highly submissive people prefer those who are dominant.

que stions

KEN BOLLEN Henry Rudolph Immerwahr Distinguished Professor, Department of Sociology Adjunct Professor, Department of Statistics Director, Odum Institute for Research in Social Science at UNC-Chapel Hill As part of SSRI’s Monday Seminar Series, the professor of sociology and statistics recently visited Duke to present his lecture on “Fixed and Random Effects in Panel Data using Structural Equation Models (SEMs).” He gives answers to some questions about his work.

Q: Much of your methodological work is with latent variable modeling. Can you explain to us non-statisticians what that means?

A: It sounds very mysterious when put in that way, but it’s basically using statistical models when you know that the things that you are dealing with cannot be measured without error. You want to do analyses that take into account that there is measurement error when you are looking at things such as democracy or characteristics of people. In your modeling, you don’t want that measurement error to give you false estimates of relationship between variables; the statistical methodology I work with addresses this.

Q: What application does this have to the work of other social scientists?

A: In virtually any problem that a social scientist is examining, when it comes to doing a quantitative analysis, the type of variables that people have in their data sets are often imperfect representations of what they would actually like to look at. So in virtually any quantitative analysis, these techniques can be used to take account of the measurement error that’s in the variables and may be obscuring the relationships that they are interested in.

Q: What are the areas of your substantive research?

A: The longest running is the problem of democratization in countries. The focus is on political systems and what is sometimes called liberal democracy: is it a political system that permits open and fair elections with some competition?

Is it a system that permits the media to openly criticize the government or others? Does it permit groups to freely organize? We’re first looking at whether we can measure it in some reasonable way, and then, if we can take care of the measurement error, what type of factors might explain why one country is democratic and another is not. I also work in the population area; most recently my work there has focused on trying to understand the role of socioeconomic status in influencing things such as the fertility of women in developing countries.

w w w. ss r i . d u k e . e d u 1 9

UNC News Service

Q: And why is this important?

A: The statistician or the quantitative social scientist can test ideas to see if what is suspected to matter really does matter. Statistics, in some ways, is a tool—just like a microscope is a tool for certain biologists studying things that aren’t visible to the human eye.

Statistics and the social sciences help us see relationships between variables when there are multiple variables that might influence an outcome. For example, what are the factors that contribute to someone being in poverty? How can we predict who gets AIDS and who doesn’t? To really examine several factors at once and try to determine if one seems to dominate over the others, you need to have some tools to enable you to do that. Statistical methodologies can help pull apart these separate variables.

Q: What got you involved in this type of research?

A: In graduate school, one of my areas of concentration was in quantitative methodology in sociology. What pushed me was the substantive work I was doing: the democratization of nations. As I tried to do an empirical analysis of that, I became less and less content with the type of measures of democracy that existed and became more interested in how you could measure it better. And if you couldn’t measure it better, how could you take into account the measurement problems when doing your analyses. And as I learned these latent variable techniques, I got deeper and deeper into the methodology itself. It certainly was driven by my interest in the substantive area.

tech nology

by Mary-Russell Robe rson

Coming to a Desktop Near You: High Performance Social Science Data Analysis

Five, 64-bit nodes with dual or quad processors and eight gigs of RAM per node. If that doesn’t mean much to you, don’t despair.

All you need to know is this: beginning in 2008, if you’re a Duke graduate student or faculty member in the social sciences, highspeed data analysis is as close as your desktop.

The Social Science Research Institute (SSRI) has installed a cluster of high-performance computers devoted solely to social science data analysis. Users can access the cluster from any computer, anywhere. The system is similar to those at Cornell and N.C. State universities. “Rather than having to spend money putting a 64-bit computer on your desk for analysis, you can access a remote 64-bit machine with a large amount of memory,” says Neil Prentice, Director of Social Sciences IT Services for Arts and Sciences Information Science and Technology.

Users should find that their analyses run at least as twice as fast than on an average desktop. “We’ve been beta testing it for six months, and all the beta testers have been ecstatic about it,” says Chris Kerr, IT manager for SSRI. “It has cut their analysis time in half in some cases.”

The cluster has five nodes, which means the work is divided among five processors or machines, Prentice says. The system includes advanced versions of social science analysis software, such as Stata, that can recognize the multiple processors. “This particular flavor of Stata allows the program to use multiple processors so it can actually spread the job across a couple of computer brains

20 gist from the mill • Spri ng 2008

instead of a single one, so jobs run faster,” Prentice says.

In addition, the machines themselves possess more memory than the typical desktop, so handling large data sets is no problem. The machines all have dual or quad processors, with eight gigabytes of RAM per machine.

All that memory allowed Brendan Nyhan, a Ph.D. candidate in political science, to interact normally with some very large data sets for the first time. For his dissertation on presidential scandal, Nyhan is using a data set that includes a monthly entry for every member of Congress for a 20-year period; his desktop computer could barely open it. “It’s hard to do research and have a flow of ideas when you’re sitting twiddling your thumbs for 15 minutes waiting for things to run,” Nyhan says. “Running it on the cluster makes a huge difference.”

Users with a valid log-in ID can access the service from any computer in the world. Site licenses for the software offered on the cluster allow up to 10 people to use the service at once. “The five computers behave in concert, allowing multiple people to log on at any given time, so one person won’t monopolize all the resources on any machine,” Prentice says. “If the cluster continues to be utilized, then we may end up dedicating more hardware to it in the future.”

Jacob Montgomery, a graduate student in political science, recently used the cluster as a research assistant on a project led

by Professors Scott de Marchi and Jay Hamilton to study factors associated with voter turnout.

Social Science @ Duke Social Science Departments African & African American Studies www.aas.duke.edu/aaas Cultural Anthropology ca-www.aas.duke.edu Economics www. econ.duke.edu Program in Education www.duke.edu/web/education History www-history.aas.duke.edu Political Science www.poli.duke.edu Psychology & Neuroscience pn.aas.duke.edu

University Institutes and Centers (UIC) Duke Global Health Institute globalhealth.duke.edu Duke Institute for Brain Sciences dukebrain.duke.edu Institute for Genome Sciences & Policy www.genome.duke.edu John Hope Franklin Humanities Institute www.jhfc.duke.edu/fhi The Kenan Institute for Ethics kenan.ethics.duke.edu Nicholas Institute for Environmental Policy Solutions www.nicholas.duke.edu/institute Social Science Research Institute www.ssri.duke.edu

Public Policy pubpol.duke.edu Sociology www.soc.duke.edu Program in Women’s Studies www.duke.edu/womstud w w w. ss r i . d u k e . e d u

21

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 f ro m t h e M i l l

upcoming events March Notes on a Scandal 7:00 p.m.

Bryan Center, Griffith Film Theater

4

fvd.aas.duke.edu/screensociety/ethics2008.php

6 Gene-environment Interactions in Mental Health 4:00 p.m.

Bryan Research Building, Room 103 genome.duke.edu/education/seminars/genomes4

19 Muppet Diplomacy: How Sesame Street is Changing our World 5:30-6:30 p.m.

Sanford Institute, Room 04 childandfamilypolicy.duke.edu

25 Workshop on Qualitative Software Comparison 3:30-5:00 p.m.

Erwin Mill Building, Room B140 www.ssri.duke.edu/training.php

26 Remembering Past Atrocity:

24 The Difference Time Makes:

Monuments, Memorials and Museums in Comparative Perspective Noon-1:00 p.m.

John Hope Franklin Center, Room 240

Latent Growth Curve Models of Women’s Political Representation 5:30-7:00 p.m.

Perkins Library, Breedlove Room

jhfc.duke.edu/fhi/events/wednesdays/index.php

poli.duke.edu/newsevents/

APRIL Legitimating Beliefs: Concepts, Indicators, and Evidence 5:30-7:00 p.m.

John Hope Franklin Center, Room 240

3

jhfc.duke.edu/ducis/GlobalEquity/schedule.htm

9 Muslim Women in Film— “Love for Share” 7:00 p.m.

Bryan Center, Griffith Film Theater fvd.aas.duke.edu/screensociety/schedule.php

12 Moral Mathematics: The Science of Modern Human Rights 8:30 a.m.

Old Chem 116 duke.edu/web/rightsatduke/calendar.html#math

25 Corpus Delicti: A SeventeenthCentury German University Debates Witchcraft, Poisoning and the Law 4:00-6:00 p.m.

National Humanities Center fds.duke.edu/db/aas/History/faculty/ news_fac_archives.html?keywords=&yr=2007

MAY High School Illicit Drug Use Onset and Early School Dropout: A Survival Analysis using Fast Track Data Noon- 2:30 p.m.

Erwin Mill Building, Bay C Conference Room

9

childandfamilypolicy.duke.edu