Researchers need more data to show what works and why

Print More

Teachers are paid more for additional years of service and additional degrees, but neither of those factors leads to better student outcomes, researchers have found. Ditto for teacher certification.

Teachers are paid more for additional years of service and additional degrees, but neither of those factors leads to better student outcomes, researchers have found. Ditto for teacher certification.

Now, with data systems following children from kindergarten to college and into employment, researchers have a lot more information and are hoping to pinpoint why some teachers are successful and others aren’t.

Studies already have shown that regardless of the school – good or bad, rich or poor — good teachers can push students up a full grade level and bad ones can hold him down a grade level, Jane Hannaway, the founding director of the Education Policy Center at the Urban Institute, told colleagues Thursday at a seminar on outcome measurements across the education, child welfare and health systems.

“In the worst schools, you find stars doing terrific things,” she said.

The education research agenda also includes figuring out the effect that principals have on student outcomes. Teachers and others stress that principals do make a difference, but Hannaway says that so far there is little empirical evidence to show that.

She’s also involved in examining the effectiveness of teacher incentive programs with these data systems.

Across the board, panelists acknowledged shortcomings in their research methods:

• Hannaway noted that researchers’ reliance on test scores can short shrift learning that is not measured by tests.

• Alyna Chien, a health science researcher at Children’s Hospital in Boston, said that the evaluation of children’s health care looks more at how many times they go to the doctor and whether they are treated with appropriate medicines than whether the child was at grade level or how often he is absent because of asthma or diabetes.

• Fred Wulczyn, a research fellow at Chapin Hall, said that the child welfare system looks more at the number of successful placements and whether a child is safe than the quality of his or her life.

 

For the most part, the data in these systems are not being meshed. But when they are, Wulczyn said, they can yield fruitful information. For example, Chapin Hall was able to look at school performance of children before and after they entered the child welfare system. They found that children were generally behind grade level before they came into the system.

“It makes us ask, How do we produce gains for these children as we take responsibility for them?’” Wulczyn said.

The seminar was co-sponsored by the Urban Institute and Chapin Hall.