Race, Ethnicity, and Academic Achievement - Proposal of Research Design This research will study the relationship between race and ethnicity and academic achievement.

This study will look at whether students from five racial classifications - White, Asian, Black, Hispanic and Native American - show statistically significant differences in their scores in reading and math. Using disaggregated data, this study will also examine whether significant differences occur within these racial classifications. In addition, this paper will study whether socio-economic status and language proficiency have a significant effect on a child's reading and math skills.

Because of many socio-economic factors linked to the construction of race, this paper predicts that white and Asian students will show higher test scores in reading and math. However, because racial classifications can obscure the differences between ethnicities, this paper also predicts that there will be significant differences in test scores within the racial categories themselves. In addition, affluence is predicted to affect the scores positively while a lack of language proficiency is expected to lower the scores.

Purpose

This goal of this paper is to contribute to studies about racial classifications and educational policy. If the test scores of vary significantly between the groups, measures should be taken to address the problem and ensure that the educational needs of all children are met.

In addition, by examining the disaggregated data, this paper will examine whether the broad racial categories are an accurate method to classify students in educational policy research. These broad racial categories may mute important differences among racial subgroups. Again, if such differences occur, they should also be addressed through educational policy.

Finally, while the primary goal of this study will be to analyze the effect of race and ethnicity on a student's reading and math skills, this paper will also look at how differences both among and within racial classifications may be affected by socioeconomic status and language proficiency.

Review of Related Literature

Much of the scholarly literature on the relationship between race/ethnicity and academic achievement focused on a Black/White dichotomy. This was because older classifications grouped students who were not white as "minorities," and majority of these minority students were African-American.

In The Black-White Test Score Gap, Christopher Jencks and Meredith Phillips edit a collection of essays that reflect on the widening disparity of reading, math and vocabulary test scores of black and white students. The bulk of the essays locate the problem in socio-economic conditions -- such as poverty and lack of opportunity -- which severely curtail the early cognitive development of black children. In the introduction to this book, the authors argue that addressing educational inequity would do more to promote racial equality in other areas as well than later measures like affirmative action (Jencks and Phillips, 1998).

In Learning While Black: Creating Educational Excellence for African-American Children, author Janice Hale studies why black students are not being educated as well as white students. Taking a different approach from Jencks and Phillips, Hale looks beyond socioeconomic status, family structure of poorly trained teachers. Instead, she criticizes the "racial profiling" that occurs in many educational institutions. Thus, many Black children are put at a disadvantage because of low expectations and indifference on the part of educators themselves. To address this, Hale calls for a redefinition of the school as a family and a village, a strategy which makes education the task of educators, parents, churches, civic groups and concerned citizens, of what Dr. Martin Luther King described as the "beloved community" (Hale 2001).

John U. Ogbu and Astrid Davis' controversial Black American Students in an Affluent Suburb further challenges the thesis that racism and economic status are the main determinants of lower test scores among African-Americans. Ogbu and Davis theorized that these children of lawyers, doctors and other professionals look towards entertainers like rappers as their role models. The authors also helped popularize the belief that hardworking black children are likely to be put down by their peers for "acting white" (Ogbu and Davis, 2003).

In their study of the effect of drug use and race on school performance, Arthur Lee Dozier and Michael James Barnes compared student achievements in the Metropolitan Achievement Test, which tested students from Grades 9 through 12. They also conducted surveys and interviews with a sampling of students, regarding their identified ethnicity and drug use. Though the study focused on drug use, it found interesting results that can be ascribed to racial background. As expected, their data indicated that students without a history of drug use had higher grades than drug users, whether or not students were White or Latino. However, there was little difference in the test scores of Black users and non-users (Dozier and Barnes 1997).

Wang (1998) examines the standardized tests, focusing on whether the opportunity to study the test topics affected the test results. Wang found that Black and Hispanic students received less exposure and quality education than their White and Asian counterparts. These findings imply that the test scores of Black and Hispanic students could be significantly improved through two ways. First, Black and Hispanic students should receive more instruction or review of subjects covered in these tests. Second, standardized tests can be more tailored towards Black and Hispanic students (Wang 1998).
Methods

Data

The data for this study will be culled from the NELS '88, a nationally-representative study conducted in 1988 by the National Center for Educational Statistics (NCES). NELS '88 included students from both public and private schools. The study first looked at academic test scores of 25,000 students, and conducted four follow-ups since the initial study, the last being taken in 2000. The availability of regular follow-up studies will also allow this researcher to monitor if the conclusions will continue to hold true beyond high school and college as well.

In addition to test scores, NELS '88 also contains information regarding a student's race, socio-economic status and proficiency in English. NELS '88 also differs from other standardized tests such as the Standardized Aptitude Test (SAT) and state-based tests like California's Stanford 9. Unlike other tests, NELS offers disaggregated racial data. This allows researchers to study whether test scores vary within racial groups as well.

This study will thus use "academic achievement" as the dependent variable. Academic achievement will be quantified as the math and reading scores of eighth grade students as reported in the NELS '88. These values reflect the mean scores of students from each racial classification.

The primary independent variable of interest is race - divided into White, Asian, Hispanic, African-American and Native American. Controls will also be added for socio-economic status and language proficiency in English.

For the disaggregated studies, the variable "race" will be further divided into subgroups for which data is available. For the category "Hispanic," the sub-groups will include the following ethnicities: Cuban, Mexican, Puerto Rican and Other Hispanic groups.

For the category "Asian," the sub-groups include the following: Chinese, Filipino, Japanese, Korean, other Southeast Asian, Pacific Islander, South Asian, West Asian, Middle Eastern and Other Asian groups.

Methods

This study will use the multiple regression method for two reasons. First, multiple regression will allow the quantitative statistical study of nominal values like race, ethnicity and language proficiency. Second, multiple regression also enables the researcher to measure how factors like socio-economic status and language proficiency affects the initial figures showing the effect of race on academic achievement.

All regressions will be compared to a baseline showing measuring the how the mean reading skills (dependent variable) and math scores (dependent variable) are affected by the following independent variables: race, then a combination of race and socioeconomic status, then a combination of race and proficiency in English.

Coding will be done as follows:

Reading skills (dependent variable) - mean scores as reported by NELS '88

Math skills (dependent variable) - mean scores as reported by NELS '88

Race (independent variable) -- individual regressions, with values assigned as follows:

Asian - 1 for positive, 0 for negative

African-American -- 1 for positive, 0 for negative

Hispanic -- 1 for positive, 0 for negative

White -- 1 for positive, 0 for negative

Native American -- 1 for positive, 0 for negative

Socio-economic Status -- mean values as reported by NELS '88 independent variable)

English Proficiency -- 1 for positive, 0 for negative independent variable)

The first set of regressions will measure the effect of the various racial categories (independent variable) on reading skills (dependent variable) and on math skills (dependent variable). This will yield the basic values of how being Asian, African-American, Hispanic, White or Native American could affect a student's skills in reading and math.

The next set of regressions will add the independent variable of socio-economic status. This will indicate if being wealthy, middle class or poor mutes or enhances the effect of race on a student's reading and math skills.

The third set of regressions will then add the independent variable of language proficiency to the first set of measurements, to measure how a lack of proficiency in English affects reading and math skills.

The fourth set of regressions will follow the pattern of the first set. Reading skills and math skills will serve as the dependent variables. However, the racial category of…