Two studies unveiled last month paint a fuller picture of the persistent achievement gap in mathematics that separates the nation鈥檚 black and Latino students from their higher-achieving white peers, and the kinds of efforts that might鈥攐r might not鈥攈elp to eliminate it.
Since at least the 1970s, federal statistics have shown that some groups of minority students tend to lag behind white and Asian-American students academically, but experts differ on the primary causes of those gaps or on how to narrow them. The new studies鈥攐ne by current and former researchers from the RAND Corp., the Santa Monica, Calif.-based think tank, and the other led by researchers from the Princeton, N.J.-based 91制片厂视频al Testing Service鈥攄on鈥檛 pinpoint any new cures either. But both call for broadening the approaches commonly used to address the problem now.
is available online from . () The ETS report, 鈥淭he Black-White Achievement Gap: Do State Policies Matter?,鈥� is not yet completed.
鈥淥ftentimes, educational researchers focus on education policy in thinking about the achievement gap, but we also need to think about extending that to welfare policy,鈥� said Mark Berends of Vanderbilt University, the lead author of the RAND study, 鈥渁nd to think about extending opportunities not only to students but to parents of students in schools.鈥�
His study draws on four different federal databases to analyze the achievement-gap trends among high school seniors between 1972 and 1992, a period when black and Latino students seemed to be catching up with white students. The studies that RAND used were: the National Longitudinal Study of the High School Class of 1972; the High School and Beyond Study; the National 91制片厂视频 Longitudinal Study of 1998; and the congressionally mandated National Assessment of 91制片厂视频al Progress.
While the tests taken by 12th graders in all those studies differed, the researchers were able to extract items that tested common mathematical concepts.
Based on those questions, the RAND authors calculated that the black-white test-score gap in math shrank by 20 percent between 1972 and 1992. Between Latino and white 12th graders, achievement differences decreased by 32 percent over the same period.
Social Conditions, Tracking
More important, the authors found, improvements in social conditions for black and Latino families tracked closely with those reductions. They calculate that 54 percent of the decrease seen in the black-white gap over that period could be accounted for by gains in the income and educational levels of black families and in black parents鈥� occupational status. Likewise, Latino students scores鈥� rose as their families became more prosperous and better educated from 1972 to 1982, but leveled off or decreased over the next decade as conditions worsened for them.
Decreases in the size of the gaps within schools also correlated with increases in the number of black and Latino students taking academic-track courses rather than general or vocational studies. For instance, those coursetaking improvements were linked to 60 percent of the decrease seen in the achievement gap between white and black students within schools, the report says.
On the other hand, academic disparities increased as schools became more racially isolated. Students attending predominantly black or Latino schools tended to do worse than students from the same racial or ethnic groups at schools with more diverse enrollments.
That鈥檚 important, the RAND report notes, because statistics suggest that schools are becoming more segregated.
鈥淚f the trends that we hear in the larger literature and from our own analysis hold up, that may be an explanation for the persistent inequality we see in black-white test scores over the last ten years,鈥� said Mr. Berends, an associate professor of public policy and education at Vanderbilt, which is located in Nashville.
In contrast to the improvements from 1972 to 1992, the period from 1992 to 2002 was marked by 鈥減ervasive, profound, and persistent鈥� achievement gaps, according to the ETS study, which was presented last month at the annual meeting of the American 91制片厂视频al Research Association in Montreal.
State Policy Effects
In looking at that decade, researchers from the testing service drew on data from 8th grade NAEP tests in math and linked the results to the education policies states employed at the time. The 10 states they studied are: California, Kentucky, Maryland, Michigan, New York, North Carolina, South Carolina, Tennessee, Texas, and Virginia.
Some states focused, for example, on setting higher academic standards and developing tougher curricula. Others took steps to raise teacher quality or to tie rewards and consequences to changes in students鈥� test scores. Some worked to reduce funding inequities among districts.
No one policy approach seemed to hold the key to reducing the achievement gap between black and white 8th graders over that decade. But states that scored high or in the middle category for their across-the-board efforts, such as New York and North Carolina, also tended to be most successful at improving black students鈥� achievement. Good overall policy ratings also linked, to a lesser degree, to progress in decreasing achievement disparities in math.
鈥淚t shows that policy is a blunt tool, but a tool nonetheless,鈥� said Henry I. Braun, the study鈥檚 lead author and a distinguished presidential appointee at the ETS. 鈥淎nd we need to focus on all policy levers.鈥�
The study also found that black-white gaps within different economic strata in states were as large as they were for each of those states as a whole. In other words, for those 10 states at least, the gaps were not confined to the poorest families.
Promising Strategies?
鈥淚t doesn鈥檛 by itself lead us to a clear picture of what we should do in other states to close the achievement gap,鈥� said Lauress L. Wise, the president of Human Resources Research Organization, an Alexandria, Va.-based research group. He reviewed the ETS study.
But, with data for more states, and for the districts within them, such analyses could suggest some promising strategies, Mr. Wise said.
