All across the nation, school districts are under pressure to raise the quality of their teachers by training them better or monitoring them more closely.
They might be better off just giving them a chance to talk to each other, says Carrie Leana, the Gordon H. Love professor of organizations and management at the University of Pittsburgh's Katz Graduate School of Business.
In an award-winning study of the Pittsburgh Public Schools, Dr. Leana found that in the schools where teachers talked to each other the most about their jobs, and where the principals did the best job of staying in touch with the community, students had noticeably higher reading and math test scores.
Even more significant was the discovery that these communication networks had a much bigger impact on test scores than the experience or credentials of the staff did.
The economic phrase for these relationships among teachers and between principals and the outside community is "social capital," while measurements of experience and training are called "human capital."
And the bottom line of her study, Dr. Leana said, is that social capital "is a more powerful predictor" of success than human capital, and "if you had to invest in something, you'd be better off investing in social capital than human capital."
Unfortunately, she said, most school systems don't realize that and put far more energy into formally training teachers or evaluating their performance than they do in providing opportunities for them to mentor each other.
It's not just a problem in education, she added. The same misplaced priorities afflict many other businesses and nonprofit organizations.
"I think we all intuitively know that our relationships at work matter," Dr. Leana said.
"I think valuing that, even in nonprofit settings like schools, seems to be difficult [for managers], and part of the reason for that is that the gains are long-term and in schools, just like in corporations, they need the gains to be 'now.' An administration will be judged on how much student achievement scores improve this year, and not five years down the road."
Dr. Leana, who grew up in Bellevue and came to Pitt 20 years ago from the University of Florida, also has studied social capital in other fields.
In one study, she and her colleagues found that low-paid child care workers who did the best job of collaborating with each other to adapt their work to changing situations provided the highest-quality care.
Another study that is still under way will follow 1,500 nursing aides over the next two years as they work in nursing homes, private homes and other settings.
And she already has reached one tentative conclusion.
Nursing home residents spend about 90 percent of their time with nursing aides, she said, and yet even the most committed and compassionate aides are not usually included in the meetings that are held to discuss the care of the residents.
"Their great frustration is that the administrators and families don't realize how much they know about the patients. Several of them have already mentioned, 'Why aren't we in the case meetings about these patients? We're there with them over long periods of time, yet they don't value our opinion in these assessments.' "
In the Pittsburgh schools study, and another similar one that was just completed in New York City, Dr. Leana said she and her fellow researchers started with one basic fact -- "most of student achievement is what walks through the door."
In other words, a student's family background and poverty level will have the biggest single impact on his test scores, so that from the outset, teachers have only a limited ability to improve student achievement.
And that's a formula for high stress, she said.
"If I were a teacher in today's environment, I would feel so threatened, because here you're handing me a job where most of the variance of how these kids are going to do is determined when they walk in the door, so I have a limited range of influence, and yet I'm responsible for the whole pie."
In the Pittsburgh study, done between 2000 and 2002, the researchers surveyed and observed teachers and principals in 88 schools, and found that social capital "explained 9 percent of variance in math scores and 18 percent in reading scores, and that's huge."
In many districts, there has been a push for principals to monitor teacher performance more carefully.
But the local study found that the principals whose students did best were the ones who invested more energy in talking with people outside the school, such as parents who could volunteer, foundations that could fund special programs, or other principals who could discuss what techniques worked best in their schools.
Inside the school, she said, the best thing principals could do is "encourage interaction and connections among the faculty."
If teachers from the same grades could get together for a half hour to an hour each week just to talk about classroom challenges together, it would probably have more benefit than all the special in-service days wrapped together, she said.
Especially in elementary schools, "teachers have no time together during the day. Every minute is taken up either in the classroom or ... monitoring lunch, and if they have any extra time, they find some cubbyhole and spend it on grading papers or getting ready for the next day."
The value of teacher-to-teacher communication was especially apparent in the New York City study, which looked specifically at math instruction in elementary classes.
In general, she said, "elementary teachers don't like math and are underconfident in their ability to teach it."
So if they have concerns or questions, she said, they aren't likely to go to their principals or special math coaches, because it would be "an admission of deficiency."
"To go to someone and say, 'I have fractions coming up next week; what do I do?' requires some trust in the other person, so you go to other teachers," she said.
The study, which has not yet been published, found that the best results came in schools where teachers not only could find time to talk with each other about math instruction, but where at least a few teachers had strong math skills.
The Pittsburgh study, which was published last year, won the Alfred P. Sloan Foundation national prize this year for best paper. When she was asked what school administrators had done to follow up on the study's results, though, Dr. Leana minced no words.
"Nothing," she said.
She speculated part of the reason may be that educators often don't pay much attention to research unless it's done by other educators.
And in her opinion, educational research is frequently off-base.
In educational research, "what you see is this tremendous bias toward looking at things like curriculum, but I sort of agree with [New York City Schools Chancellor] Joel Klein that the problem with public schools isn't that we haven't come up with the best curriculum. The problem with public schools is that they're so poorly managed."
Mr. Klein, a former Justice Department lawyer, "has a wonderful story about how there was a leaky roof in one of the schools, and they kept sending someone to fix the floor, because the dripping water was damaging it, and they fixed the floor three times, but they never fixed the roof."
When school districts spend too much time on things like selecting a curriculum, and not enough time in promoting mentoring opportunities among teachers, she said, "I think that's an example of not fixing the roof."
Mark Roth can be reached at firstname.lastname@example.org or at 412-263-1130 First Published November 26, 2007 5:00 AM