NCES Looks to Bridge Longitudinal School Information
The National Center for Education Statistics is looking for ways to plug holes in our understanding of children's trajectory from elementary to secondary school.
NCES, part of the Education Department's research agency, has conducted more than a half-dozen major longitudinal studies over the years, but stop short of painting a complete picture of a student's educational path.
"One thing that has always frustrated me is, for all these studies we do, somehow they've never lined up to allow us to study a cohort through their entire education and beyond," said NCES Commissioner Sean P. "Jack" Buckley. "It is practically impossible to take a sample like that."
Because any long-term study naturally loses some students through the years to anything from school transfers, family moves or even death, NCES must recruit thousands of students in the initial cohort to keep a sustainable pool of students to study. Yet Buckley said the percentage of schools willing to participate in such studies—which often require additional testing and paperwork for schools—has dropped from 70 percent or more during studies in the 1970s and 1980s to only 50 percent in the recent 2009 High School Longitudinal Study.
Moreover, it has left some data dead zones in critical periods of schooling. "Middle school is an area we have completely not studied, except one early childhood cohort," Buckley said. "We have never captured an entire trajectory through middle school."
NCES hopes to fill that gap with a bridge study connecting data from two ongoing longitudinal studies of kindergarten through grade 5 cohorts and high school cohorts. Researchers are using retrospective administrator data, in part from state longitudinal databases, to synchronize the two pools of students. In prior studies, NCES has found that connecting student cohorts using administrative data builds as useful a picture of the group as enrolling 20 percent of the students in a continuation study between the two grade-level studies—and at lower cost.
So far the 2009 High School Longitudinal Study is including representative samples of middle school data from California, Texas, Florida, Georgia, Michigan, North Carolina, Ohio, Pennsylvania, Tennessee and Washington, and NCES is conducting simulations to build a better picture of these students' transitions from elementary, through middle school, and into high school.
Peggy McCardle, the chief of Child Development and Behavior Branch of the National Institute of Child Health and Human Development, praised the project at a presentation last week. "We can't not do this pre-adolescent period; this middle period is far more important than I think any of us realized," she said.
McCardle noted that NICHD frequently relies on NCES data to answer policy questions, such as when girls tend to lose interest in science. "We use these data sets all the time, but we need that middle school piece," she said. "We know girls turn off science in 4th to 8th grade, but if you don't get this cohort in there, you're not going to answer the question of why that happens and what you can do about it."