Volume 6, Number 2, 1997


School choice study may answer key questions

A major, three-year study by faculty in UD's College of Education will analyze an untapped national database to answer some key questions in the school choice debate-a hot topic in education today.

Doug Archbald, David Kaplan and Yas Nakib, with graduate student Jeanine Molock, Delaware '02M, will conduct the research, funded by a $420,000 grant from the Office of Educational Research and Improvement in the U.S. Department of Education.

According to Archbald, the study will improve the public's understanding of policies and outcomes of school choice, as well as inform education leaders and improve policy analysis on a national level.

Their study will merge for the first time demographic data from the U.S. Census Bureau, achievement data based on national tests, school district student enrollment and resource-use data, plus survey information on school choice and other types of student assignment policies from a national sample of school districts.

The data will answer such important questions such as:

Archbald initiated the grant and is the project's school choice expert. He says school choice policy has grown dramatically over the past 15 years, although its roots go back much further.

He began his research on school choice policy 12 years ago when he was a researcher in a large project in Milwaukee. He studied magnet schools there (schools that offer special programs to attract students from diverse backgrounds interested in a particular subject matter). Magnet schools were one of the first large forms of school choice, and Milwaukee one of the places leading the way.

The Milwaukee study led not only to Archbald's dissertation but also to his participation in a national study funded by the U.S. Department of Education to evaluate the growth of magnet schools and programs and their use of federal funds.

Interestingly, that study and others have shown that school choice programs are commonly implemented after the rescinding of desegregation orders.

"School choice brings about new enrollment patterns, that is, how kids are allocated among schools. Years ago, there were basically just neighborhood schools; then, especially in the '60s and '70s, many districts started desegregation programs-redrawing school boundaries, reassigning students and using busing to achieve racial balance. School choice is another principle of allocation," Archbald says.

"With magnet schools and school choice, neighborhood boundaries become less important," he explains, "but it can't just be a free-for-all, especially in larger school systems often with 80,000 or 100,000 kids and 100-200 schools." All choice programs, he says, have to have guidelines.

School choice is a big change in student assignment policies, and there is no national study of what works and what doesn't. Equity issues, geography and racial balance all play a role in implementing school choice. How school systems have managed these concerns will be part of the study. Archbald says he believes individual school systems have chosen to solve problems one by one in many different ways.

"School choice policies can mean a dozen different things," he says. "You no longer have the notion that all schools should be the same. You may have a math/science school, a school of the arts, a charter school. Parents and students can be faced with many choices. You may have choices among the regular schools within the district, as well as a charter school or two and inter-district options."

Nakib, an education finance specialist who recently joined the UD faculty, will focus on issues related to the costs of implementing school choice policies and potential resource inequalities developing among "chosen" and "unchosen" schools.

"It is not enough to know how the allocation and use of resources in schools with and without choice differ, but also whether resource allocation practices lead to better outcomes," Nakib says.

Kaplan brings expertise in measurement and statistics. The main analytical strategy will be multilevel linear modeling.

According to Kaplan, "Effects of district choice policies on the achievement and opportunity of students cannot be accurately assessed by either aggregating data to the district level or disaggregating data to the student level. Our multilevel models will allow examining effects of district policies as they are manifested throughout the levels of the school structure and ultimately to student level outcomes of interest." The researchers estimate it will take a year just to develop the data set.

-Beth Thomas