1/2 Dean + 1/2 Dean = 1 Deanship

[an error occurred while processing this directive]

November 18, 1999

SEARCH THE GAZETTE

HARVARD GAZETTE ARCHIVES

John Willett and Judith Singer share the post of academic dean at the Graduate School of Education. Willett says: "In statistics, that's called an interaction. It's where the sum of two people is greater than the addition of their parts. So we prefer to be called a 'statistical interaction.'" Photo by Jon Chase.

Education Professor Judith Singer sits calmly, while her colleague John Willett leans back, balancing his chair on its two back legs. "Don’t do that," she jokes, "If you crack your head and die, I have to do this job alone."

Willett and Singer are pursuing a novel experiment at Harvard. This fall, the two longtime research collaborators started sharing a new project: the post of academic dean for the Graduate School of Education. Singer and Willett share responsibility for overseeing the academic life of the school, enhancing the professional development of faculty, advising on faculty searches, and ensuring that graduate students have sufficient financial and academic support.

"I was taken by the idea of having ‘the Fred Astaire and Ginger Rogers’ of the statistics world take their act to the academic deanship," says GSE Dean Jerome T. Murphy. "Their research and teaching collaboration has been extraordinarily rich and fruitful. Seeing that partnership translate into administration and reaping its benefits will be exciting for the School."

A ‘Statistical Interaction’
Although Singer and Willett officially each act as half a dean, they don’t split responsibilities by days or broad job functions, as in other job-share situations. Instead, they approach the position the same way they have handled their 15-year research collaboration. Initially they double-team on most tasks to learn the ropes, then gradually divide the work by subtasks, rather than overall functions. Eventually, each will have a hand in matters related to faculty or student affairs with no one solely governing any one function.

Their office layout reflects this relationship. Rather than locate one person in one office and another in the next – where people might naturally draw comparisons between who got the smaller or larger office – Willett and Singer have divided the offices by function, having a joint meeting space and a joint office space.

Most would call their working style synergistic. Willett says, "In statistics, that’s called an interaction. It’s where the sum of two

people is greater than the addition of their parts. So we prefer to be

called a ‘statistical interaction.’ "

Aside from student advising, they work absolutely collaboratively, sharing one set of files and a conference room where they hash out details.

A Model for Others
"The notion of distributing an academic deanship between two people is an interesting idea," Singer says. As universities struggle to fill administrative posts with faculty reluctant to take time away from teaching or research, Willett and Singer’s arrangement could provide a new model. As a team, they share goals for the advancement of their research and administrative work.

"We’re research collaborators as well as administrative collaborators," says Singer, "so we have common goals in both aspects of our lives. There’s an incentive for each of us to facilitate our careers jointly. For example, now that we’re sharing the deanship, we’re policing each other and saying, ‘You’re not supposed to be here, this is your research time!’ "

Singer and Willett agree that their co-academic deanship could provide a model for other schools. But, they argue, to work effectively and efficiently together in administration, two faculty members should have past collaborative experience with each other. Two faculty members randomly chosen to share an administrative post would have a difficult time learning to work in tandem. Universities should court faculty who have already collaborated on research, teaching, or advising.

One myth they feel their experience is putting to rest is that collaborations are harmonious pairings. "There are probably no two people in the school who disagree with each other more than me and Judy," Willett says.

"When people think about collaborators they often have the sense that collaborators agree on everything, and that the way in which you develop consensus is from sharing a vision," says Singer.

"But it takes more than a shared vision of an ideal ‘end.’ The strategies or ‘means’ we would recommend to get to this ‘end’ would be quite different. Since you respect the other person’s opinion, if the other person disagrees with you there must be some value in that point, and that point must be considered and dealt with if one is to come to a true consensus."

The key, Willett and Singer say, is to keep disagreements private and show a common face to the world by going public with decisions only after they’ve built consensus between themselves.

"It’s similar to the way parents deal with kids: they may argue over the best way to discipline the child, but in the end they show the child only one face," says Singer. "Except, of course, that we’re not married," laughs Willett.

How It All Started
Their partnership began with a mailbox. When John Willett arrived at the Graduate School of Education, he found his mailbox stuffed with pink message slips – all from Judith Singer, the other statistician hired to teach at GSE. It was 1985, and the School was unlikely to tenure two professors who specialized in quantitative research. Willett already felt he was lagging behind, having spent the fall semester finishing up his graduate work at Stanford, while Singer had arrived and settled into GSE teaching the two core statistics courses.

Willett and Singer met for lunch and found much in common and soon began talking statistics and co-authoring research papers. Then Singer gave Willett a surprise.

"Judy just handed me one of the classes at lunch one day," Willett explains. "I might not be here today but for that act because I may not have developed a constituency. In doing that, she forged this bond between us."

Eventually, the two navigated a successful joint tenure process with the help of senior colleagues and mentors. Their research has garnered a host of honors, including the prestigious American Educational Research Association’s 1992 Raymond B. Cattell Early Career Award for Programmatic Research and its 1993 Review of Research Award.

Although Singer and Willett’s primary interests lie in education and the social sciences, they have found that their work has relevance to scholars collecting and analyzing longitudinal data in a number of fields. On a recent database search they conducted, the pair found that their work had been cited in hundreds of papers in everything from criminology and atmospheric science to veterinary science and forestry.

"People have been able to take what we’ve been writing primarily on humans, in school settings, and then apply it across a wide range of disciplines," says Singer.

The two recently were invited to participate in a conference on anorexia and treatment modalities for anorexia and bulimia.

"We thought, ‘What do we know about that?’ But it turns out that where we’ve been approaching our work from the perspective of ‘How long does a kid stay in school,’ they’ve been approaching their work from ‘How long does an anorexic take to recover? ’" Singer explains. Psychiatrists and doctors need to follow their patients’ progress with particular treatments over time just as education reformers follow the effects of reform movements on districts, schools, and individual children.

With support from the Spencer Foundation, Singer and Willett are finishing a book on the analysis of longitudinal data, Twice Is Not Enough. Also in the future is a companion book (with the working title The Best Laid Plans) describing recent advances in the design of longitudinal research.

Willett and Singer have cowritten two books, By Design: Planning Better Research in Higher Education (with R. Light, 1990) and Who Will Teach? Policies That Matter (with R. Murnane, 1991), as well as dozens of articles on statistical methodology, research design, and education policy.

With common research, administrative duties, office space, and career goals, Willett and Singer share just about everything in their professional lives, but they do draw some lines in the sand. "We tried to share the same keyboard once," Willett jokes, "but we were harrassing each other mercilessly, so we threw that idea out the window."