Graduate Record Examination

Every year, thousands of students prepare for, and take, the Graduate Record Examination (GRE)–a standardized test that measures the aptitude of promising graduate students. In 1998, 364,554 potential graduate students, a number that includes one-third of all bachelor degree recipients, took the GRE, according to the National Center for Education Statistics. Why does this test draw thousands of students? GRE scores are required for acceptance into most graduate programs. As a result, the exam, designed to assist graduate schools with their admissions decisions, also succeeds in sending would-be graduate students into a panic, raises questions of the efficacy of standardized testing, and funds a lucrative "test-prep" industry. Despite such side effects, it has proven useful to those it intends to serve–graduate schools. The GRE provides these institutions with a universal ruler against which applicants can be measured.

Administered by the Educational Testing Service (ETS), the GRE takes three basic forms: the General Test, Subject Tests, and the Writing Assessment. The General Test, most often referred to as the GRE, measures verbal, quantitative, and analytical reasoning skills. The Subject Tests measure achievement in eight disciplines. The Writing Assessment consists of two analytical writing tasks. Graduate institutions and departments independently determine which GRE test, or combination of tests, will be required for admission.

Genesis of the GRE General Test

Created in the early 1930s, the Graduate Record Examination General Test has been assessing college student aptitude for nearly seven decades. The Carnegie Foundation for the Advancement of Teaching funded the project, unaware of the vital part the test would soon play in graduate education. Widespread use of the GRE General Test began after World War II, when colleges and universities selected the test to help in the evaluation of the larger, more diverse pool of students applying to graduate programs. From this point forward, GRE test results have been used to evaluate students' capacity to succeed in graduate programs. Today, the Educational Testing Service and the GRE Board define the test as "a measure of knowledge and skills that members of the graduate community have identified as important for graduate study" (GRE 2001a, p. 2).

General Test

The Graduate Record Examination General Test's three parts–verbal, quantitative, and analytical–are designed to measure examinees' reading comprehension, mathematical, interpretative, and logical reasoning skills. Taken in a timed, proctored setting, the multiple-choice written test causes stress for many students. Though anxiety over the test has not changed, the format has. On April 11, 1999, the option to take the exam in written form was eliminated (except in developing countries). Today, the Computer Based Test (CBT), which replaces the pencil and paper version, is offered daily in centers around the world. Soon-to-be test-takers, and GRE test preparation companies, such as Kaplan Incorporated and The Princeton Review, have had to adjust to the features of the new design. In the past, paper-based GRE questions were identical: presented and weighted equally for all test-takers. Now, the "adaptive" CBT presents students with unique combinations of questions tailored to their skill level. Students complain about early questions on the computer-based test having greater bearing on their final scores and about the inability to answer questions out of sequence. Many agree, however, that the benefits of CBT, including instant scoring, continuous availability, and better data analysis capabilities, far outweigh the disadvantages. Research indicates that scores from the paper-based tests are comparable to those of the computer-based test. Each section of the test yields a separate score ranging from 200 to 800. Scoring may change in October 2002 when one of the multiple-choice analytical sections will be replaced by a writing component.

Subject Tests

The GRE Subject Tests, offered in both paper and computer-based formats, are designed to help graduate school admission committees and fellowship sponsors assess applicants' preparation for graduate work in specific disciplines. The Subject Tests, offered in Biochemistry, Cell and Molecular Biology; Biology; Chemistry; Computer Science; Literature in English; Mathematics; Physics; and Psychology are required by select programs. Subject Test scores can range from 200 to 990.

The Writing Assessment Test

Launched in 1999, the Writing Assessment comprises two analytical writing tasks: one an opinion piece on an issue of the Educational Testing Service's choosing, the other a critique of a presented argument. The "tasks" are meant to complement each other, providing evidence of the examinees' ability to both make and analyze arguments. Like the Subject Test, this test can be taken on paper or electronically. Writing assessment scores range from 0 to 6.

Use of GRE Scores in Graduate Program Admissions Decisions

At every education level, controversy surrounds the use of standardized tests. According to the Business Council for Effective Literacy "objections [to standardized tests] tend to fall into two broad categories: their intrinsic defects and their misuse" (p. 6). At the turn of the twenty-first century national attention has turned to the Education Testing Service's GRE, with questions about GRE scores' assumed correlation with success in graduate school and issues of test fairness.

According to ETS, the Graduate Record Examinations measure general skills acquired over time. The scores have been validated for use in graduate admissions, fellowship selection, and other assessments related to graduate study. When interpreted and applied properly, GRE test scores can be helpful in admissions decisions. In the United States, master's and doctoral level admissions decisions are often based on a number of factors, including grade point averages (GPAs), transcripts from previously attended higher education institutions, GRE General Test scores, letters of recommendation, personal statements, self-reported information, and in some cases GRE Subject Tests, Writing Assessments, and personal portfolios. Problems arise when departments and colleges heavily rely on GRE scores, rather than considering all the information available on the student. As Judith Toyama states, "All too often, [GRE] scores, because they are numerical, [thus quantifiable], are given more importance" (p. 34). As a result of this pervasive misuse, the GRE is seen as a "gatekeeper" of minority and female–typically lower-scoring groups–access to graduate education. Many graduate schools use a combination of the verbal and quantitative section scores when considering applicants. How well does this composite predict academic success in graduate school? Peter Sacks, using data compiled by ETS, states, "Data from 1,000 graduate departments covering some 12,000 test-takers show that GRE scores could explain just 9 percent of the variation in grades of first year graduate students." Some studies have shown, however, that a combination of GRE scores and grade point averages can be indicative of student success.

Another controversial aspect of the GRE is that disparity in scores between ethnic groups points to the possibility of an intrinsic bias, or unfairness, in the General Test. Data analysis of GRE General Test scores reveals differences in the mean scores earned by racial, ethnic, and gender groups. According to the Graduate Record Examination Board, data from the 1999–2000 testing year shows the following: (1) white men scored at least 100 points higher, on average, than minority men and both minority and nonminority women; (2) self-classified whites, Asian/Pacific Islanders, and "others" received higher scores on the verbal and analytical sections than other ethnic groups; and (3) within ethnic groups, men scored higher than women. In response to concerns over such performance gaps, the Educational Testing Service and the GRE Board instituted practices to reduce GRE test bias. Faculty have been included in the test design, review process, and analysis of differential item functioning (DIF) when the percentage of correct answers on the same test question differs between groups. Despite these efforts, differences still occur along ethnic and minority lines. ETS stresses, however, that differences should be expected as "test results cannot be judged in isolation from the unequal outcomes produced by our educational, economic, and social systems" (2001a, p. 2).

Evaluation

Within our culturally, economically, and socially diverse population, no one measure could equitably evaluate individuals' promise. Even the GRE, a vetted, long-standing standardized test, has difficulty producing a fair measure of the diverse students that flock to test-taking centers each year. Moreover, when admissions committees make GRE scores the primary basis for their decisions, they overlook many characteristics and qualifications that are not measured on the test and are proven factors in student success, such as drive, passion, determination, and charisma. The use of GRE scores in the appraisal of students' ability in the tested areas is merited. The scores are inadequate, however, when applied as sole predictors of success. Students, scholars, and the ETS agree, the Graduate Record Examinations are valuable, as long as graduate school admissions committees exercise good judgment, and use the scores to supplement other admissions criteria.