Why Do the Achievement Gaps Exist?
A group of researchers finds that black and white family characteristics do not explain the same amount when it comes to understanding the changes in test scores during the 1970s and 1980s. General changes in U.S. families during this period appear to have had a greater effect on black children's test scores than on white children's test scores.
This reports some of the ideas and findings from the following source:
Grissmer, D., Flanagan, A., & Williamson, S. (1998). Why did the black-white score gap narrow in the 1970s and 1980s? In C. Jencks and M. Phillips (Eds.),The black-white test score gap(pp. 182-226). Washington, DC: Brookings Institution. Pp. 195-201
To see other reports that originated from this same citation, go to the bibliography.
The test-score gap between black and white U.S. students decreased between the 1970s and 1990s. The gain came from improvement in the test scores of black students primarily during the 1970s and 1980s.
There were many things that contributed to this increase in the test scores of black children. What part of this improvement in black test scores can we attribute to changes in the family?
David Grissmer and his colleagues Ann Flanagan and Stephanie Williamson try to tease out the influences that family changes had on black and white test scores during this period.
Grissmer and his coauthors begin by noting that in most analyses of test performance, family characteristics explain more than school characteristics. But, what family characteristics were important? The authors consider the following trends in the U.S. family between 1970 and 1990:
Grissmer and colleagues say that the gains to black students from better educated parents and fewer siblings offset any test score losses from higher rates of single parents and younger mothers. They also note that while all students gained, black students seemed to have gained more from these changes than white students.
The authors find the following changes occurred in black and white families between 1970 and 1990.
Figure 1 shows that the level of education increased for both black and white parents. However, the dramatic changes appear in the level of education of black parents.
Figure 2 shows that smaller black families became much more common between 1970 and 1990. As in Figure 1, we see similar changes for both black and white families. However, the increases in smaller families were more pronounced in black families. Grissmer and his colleagues say that this is important because fewer siblings is related to higher test scores for children.
Figure 3 shows that between 1970 and 1990, there was an increase in the number of younger mothers. This increase was slightly more pronounced in black families. Even though this change could be associated with a drop in test scores, the negative effect did not offset the gains made in parents' education and smaller families.
Figure 4 shows two trends. For both black and white families, there was an increase in the number of single mothers for 15&18-year-olds. There was also an increase in the number of older teens who had mothers working. Neither of these trends offsets the test score gains explained from smaller families and better educated parents.
The authors measure the increase in the average test scores of black and white students. They measure the gain, not in terms of actual points scored on the test, but in terms of the standard deviation (a standardized unit of measurement) gained over the group's initial score. This comparison is between scores of black or white children at the beginning of the National Assessment of Educational Progress testing (in the late 1960s) and the scores of the same age black and white children twenty years later.
They find the following increases for the two groups of students.
Figure 5 shows that black reading scores for 17-year-olds increased nearly an entire standard deviation between 1970 and 1990, compared with only a 0.11 standard deviation increase for white 17-year-olds across the same period.
Grissmer and colleagues then run statistical procedures to see how much of this gain is explained by the changes in families. They say that current estimates suggest that heredity accounts for about half of the variance in test scores. Subtracting the "half" explained by heredity, Grissmer and his colleagues estimate that family environment accounts for the following amount of in black student test scores.
We see similar trends in both black and white families. Why would the changes in the black families have helped black Americans so much more than white Americans?
Grissmer and his colleagues think that part of the explanation may come from unequal educational opportunity. Before the late 1960s it was much more difficult for black Americans to access quality education. To get a quality, advanced education, blacks had to overcome many obstacles.
For white Americans who faced relatively few obstacles to education, academic achievement may have depended mainly on innate characteristics. For black Americans, many things beside innate intelligence would have had a greater effect on academic achievement. So, changes in the black American family environment helped black children begin to "catch up" to their white schoolmates.
Why did both black and white test scores rise for all ages in both reading and math over the past 30 years?
Why did black students' scores rise substantially more than white scores at all ages and in all subjects?
Why did black adolescent achievement scores remain stable for cohorts entering from 1960 to 1968-70, then suddenly accelerate for cohorts entering school between 1968&1972 and 1976&1980, and then stabilize or fall in subsequent cohorts?
Why did black adolescents first gain more and then lose more than black nine-year-olds?
Why did black reading gains precede black math gains?
Why were black gains higher in the Southeast and lower in the Northeast?
Why did low-scoring students gain more in math and less in reading than higher scoring students, regardless of race?
Grissmer and his colleagues draw on National Assessment of Educational Progress (NAEP) longitudinal test score data. This data source begins in the late 1960s and continues.
To determine how trends in test scores are related to other social changes, the authors relate the NAEP data to family characteristics from the Current Population Survey and the National Education Longitudinal Survey and the National Longitudinal Survey of Youth.
The authors look at the changing test scores as the dependent variable. They examine three kinds of independent variables:
They compare black and white students' test scores from the initial NAEP data with later test score trends. They measure changes from this initial starting point in terms of the standard deviation. They then regress changes in family, education, and school trends onto the changes in black and white students' test scores.
Support for Grissmore, Flanagan and Williamson's research came from the center for Research on Educational Diversity and Excellence, supported by the U.S. Department of Education, the Danforth Foundation, and RAND.
Copyright © 2007 Learning Point Associates. All rights reserved.