More than half a century after the U.S. Supreme Court outlawed school segregation, a new Florida study shows that the racial composition of schools still matters when it comes to scores on student-achievement tests.
An abstract of is available online from the .
The study, published last month in the American Educational Research Journal, is based on analyses of test scores and other data from elementary, middle, and high schools in Florida鈥檚 67 public school districts.
All other things being equal, the researchers found, schools with high enrollments of African-American students tend to score lower on state mathematics and reading tests than integrated or mostly white schools.
Though the researchers concentrated on Florida, they said the implications of their findings have a broad reach because they come at a time when districts nationwide are being released from long-running court orders to desegregate their schools.
鈥淚t鈥檚 as though districts have decided that those patterns don鈥檛 matter anymore, when clearly our evidence suggests they do and quite decidedly,鈥� said Kathryn M. Borman, the study鈥檚 lead author and an anthropology professor at the University of South Florida in Tampa.
Tests鈥� Fairness Debated
The researchers said their findings also raise questions about potential inequities in testing programs, such as the Florida Comprehensive Assessment Test, or FCAT, that penalize and reward schools based on students鈥� scores.
Such tests are a cornerstone of the No Child Left Behind Act, the 3-year-old federal school improvement law championed by President Bush. It requires states to adopt high-stakes testing for schools that receive aid through the federal Title I program for disadvantaged students.
But, while the law compels those schools to show that test scores are improving for every racial and ethnic subgroup they enroll, it makes no special provisions for their overall demographic makeups, a practice the researchers contend is 鈥渃learly unfair鈥� to schools in which nearly all the students are African-American.
鈥淲hen you look at the schools being sanctioned, you鈥檒l find these programs are disproportionately sanctioning minority schools and schools with high concentrations of English-language learners,鈥� said Harvard University researcher Gary Orfield, who has found similar relationships between racial imbalance and lagging achievement in his own, separate studies of Boston-area schools. 鈥淭his is a pattern we鈥檙e seeing all over the place.鈥�
But advocates for high-stakes testing vigorously dispute the contention that the Florida testing system is unfair.
鈥淪o are they saying we should not have the same goals for those [racially isolated] students because those students do not do as well on the FCAT?鈥� said Ross E. Weiner, a policy director for the Education Trust, an education research and advocacy group based in Washington. 鈥淚 think this kind of research holds black kids back.鈥�
Ms. Borman and her research partners found that scores for schools with mostly black students were lower than those in other schools, even after accounting for other factors that are known to affect academic achievement, such as differences in per-pupil spending and the percentages of students from poor families. They also attempted to account for schools鈥� differences in instructional quality by factoring in teachers鈥� average years of experience and levels of education and average class sizes.
Once those factors were taken into consideration, the authors found, 36 percent of 4th graders passed the reading section of Florida鈥檚 state exam in schools where 90 percent of students are African-American. By comparison, the 4th grade passing rates for the same tests in schools with less than 15 percent black enrollment was 54 percent.
On 5th grade math tests, the study concludes, 5 percent fewer students passed the tests in mostly black schools than in mostly white schools.
The researchers found similar race-related achievement patterns for middle and high schools.
However, the researchers said that their study is limited because it is based on testing data from just the 1999-2000 school year, the second year that Florida鈥檚 testing program was in place.
Studies are under way to track those testing patterns over several more years, they said.
