Search

Areas of Expertise

 

Achievement Gaps
Literature Library

Topics

Why Do the Achievement Gaps Exist?

  1. Explanations

What Matters When Trying to Understand the Educational Achievement Gap Between Black and White Students?

Many factors contribute to racial inequality in achievement test scores. A researcher examines the interplay of race and institutional processes in the creation and perpetuation of racial inequality.

Citation:
This reports some of the ideas and findings from the following source:

Roscigno, V. J. (1998). Race and the reproduction of educational disadvantage.Social Forces, 76,1033-60.

To see other reports that originated from this same citation, go to the bibliography.

There are many factors that contribute to racial inequality, says researcher Vincent Roscigno. To understand the role that race plays in explaining the difference in black and white students' academic achievement, Roscigno points to the interplay of different institutional processes.

He looks at how race affects academic achievement by examining family and educational institutions. Roscigno believes that differences in academic achievement that initially appear to be racial can really be explained by looking at the characteristics of families and the educational institution.

In addition to race, Roscigno looks at three other factors or "clusters" of variables:

  1. Teachers and classes
  2. Characteristics of family and friends
  3. School characteristics

Roscigno uses data from the National Educational Longitudinal Study (NELS) and the Common Core of Data (CCD) in order to see what kind of effects these three factors have on student math and reading achievement.

Roscigno Examines Many Variables

What specific variables does Roscigno study to understand the achievement gap? The table below presents the variables that Roscigno examines, grouped according to race and the three factors mentioned above. (The fifth category is a statistical control category that estimates a student's probability of attending a public versus a private school.)

Factor

Variables Measuring that Factor

Race

  • Black versus white

Teachers and classes

  • High teacher expectations
  • High academic track

Characteristics of family and friends

  • Family income
  • 50+ books in the home
  • Parental education

    high school
    college
  • Family structure
    single parent
    stepparent
    number of siblings
  • Peer aversion to education

School characteristics

  • Student/teacher ratio
  • Per-pupil expenditure
  • Percent of students receiving free or reduced lunch
  • Black segregated school (>75% black)
  • White segregated school (>95% white)

Probability of attending a public school

>

Estimated by taking into account

  • family class position
  • race of student
  • region of the country

Which Variables Were Important for Explaining Differences in Students' Math and Reading Achievement?

After examining all the variables together, Roscigno finds that not all were important (statistically significant) for explaining differences in math and reading achievement.

The following table presents the specific variable, whether it was statistically significant, and whether it was positively or negatively associated with achievement.

Variables

Important for Math Achievement?

Positive or Negative Relation to Achievement

Important for Reading Achievement?

Positive or Negative Relation to Achievement

Race

yes

negative

yes

negative

Family income

yes

positive

yes

positive

50+ books in the home

yes

positive

yes

positive

Parent high school degree

yes

positive

yes

positive

Parent college degree

yes

positive

yes

positive

Single-parent household

no

no

Step parent and parent in the home

no

no

Number of siblings

no

yes

negative

Peer aversion to achievement

no

no

High teacher expectations

yes

positive

yes

positive

High academic track

yes

positive

yes

positive

Student/teacher ratio*

yes

negative

no

Per-pupil expenditure

no

no

Percent students receiving free/reduced lunch

yes

negative

yes

negative

Black segregated school

yes

negative

yes

negative

White segregated school

yes

positive

yes

positive

Probability of attending public school

yes

negative

yes

negative

*

Note: as the number of students per teacher goes up, student achievement goes down.

Roscigno found that, in his data, that some variables:

  • did not make a difference for either math or reading achievement: single-parent households, the presence of stepparents in the home, and per-pupil expenditure by school
  • made a difference only for math achievement: student-teacher ratio
  • made a difference only for reading achievement: number of siblings

All other variables Roscigno studied made a difference for both reading and math achievement.

The Bottom Line

Race is only one among many factors that affect math and reading achievement.

However, the problem is that race is also associated with many of the other variables that Roscigno takes into account. The problem is how to tell how much of the influence on achievement is due to race and how much is due to other factors, such as family and the educational institution.

Roscigno develops a model that can help untangle the relative effects of race and other variables on academic achievement. While all the variables are related, Roscigno says that the particular ways that they interact are important for understanding how patterns of inequality in the larger society are reproduced in the educational setting.

Research Design:

Research Question

How does the relationship between the institutions of family and education exacerbate racial inequalities in academic achievement?

Data

Roscigno draws from two data sets:

  1. The National Educational Longitudinal Study. This data set is a nationally representative sample of U.S. high school students and includes principal, teacher, and parent surveys in addition to information on student achievement. The sample includes information at both the school and the student levels. Roscigno was able to match information on students in this data set to school and district data included in the Common Core of Data data set.
  2. The Common Core of Data. This is the primary database used by the National Center for Education Statistics for public elementary and secondary schools.

Because of limitations of the data sets and because of the focus of his study, Roscigno omitted the following information from his analysis:

  • information on private school attendees
  • information on high school drop outs (6.4% of the sample)
  • information on races other than black and white.
Methods

Roscigno's goal was to compare the relative influence of race on academic achievement when other factors were taken into account. To do this he tested three hierarchical linear models of influence on student achievement—specifically math and reading achievement scores.

  • First, Roscigno analyzed the data to see "how much" a student's math achievement could be explained by looking only at race.
  • Second, Roscigno added measurements of family and peer influences into the model.
  • Third, Roscigno added measurements of educational institution attributes into the model

By examining the change in how much race explained as he moved from one model to the next, Roscigno was able to determine the proportion of student achievement—initially "explained" by race—that was really due to institutional processes in the family and educational institution. For instance, when family and educational factors were taken into account, race explained less and less variation in student achievement.

This method of examining the influence of race did not indicate that race was not important in understanding student achievement. But, according to Roscigno, race is simply one component in a larger "web" of institutional processes that lead to racial inequality in academic achievement.

Funding Source

Roscigno's research was supported by a grant from the American Educational Research Association (which, in turn, receives its funds from the National Science Foundation and the National Center for Education Statistics in the U.S. Department of Education).

 



Copyright © 2007 Learning Point Associates. All rights reserved.