On average, members of different racial, ethnic and economic backgrounds perform differently on standardized tests. These differences do not necessarily mean that tests are biased. Extensive research by ETS and other organizations has shown that these performance differences can be the result of a number of factors, such as variation in course-taking patterns, interests, knowledge and skills, or differential educational, economic and social systems in which everyone does not receive equal opportunity. These score differences are seen in all standardized tests.
Despite the extensive work that ETS does to ensure that the GRE® tests are as free from bias as possible, disparities in performance among underrepresented groups still exist. A review of all components of an applicant's file, in which GRE scores are considered as one piece of information among many, enables each applicant to be evaluated as fairly as possible.
Before considering any applications, we advise that institutions/programs determine the significance of GRE scores in relation to other components of an applicant's file. Considering students holistically ensures a more fair admissions process and is important to ensure that all applicants have the opportunity to present multiple aspects of their potential value to the program. If you are not able to do a full holistic file review for all applicants, you should pay particular attention to applicants who may have had experiences somewhat different from those of the traditional majority.
- Performance information for underrepresented groups can be found in the publication entitled A Snapshot of the Individuals Who Took the GRE® General Test.
- For more information about ETS's extensive efforts to ensure that the GRE tests are as free from bias as possible, visit Fairness and Validity.
- For more information on ETS's policy work to reduce achievement gaps, visit our Achievement Gap website.
Do Score Differences Mean that the GRE Tests Are Biased?
No, differences between various groups, which are seen in all standardized tests, do not imply bias. There are a number of factors that contribute to observed differences in scores, such as variation in course-taking patterns, interests, knowledge, and skills, or differential educational, economic, and social systems in which everyone does not receive equal opportunity.
ETS has long placed great importance on the issue of fairness. Rigorous processes to ensure test fairness are carried out by a diverse team of assessment specialists in accordance with standards set by the American Educational Research Association (AERA), the National Council on Measurement Education (NCME), and the American Psychological Association (APA).
ETS has taken significant measures to ensure the test upholds the highest standards for fairness by incorporating reviews and checkpoints throughout the development process:
- In the early concept phase, a variety of test question types were pilot-tested and any question types that tended to produce group differences in performance were eliminated from the test plan.
- During the development phase and continuing today, test questions are further scrutinized by specially trained fairness reviewers to ensure they meet rigorous standards. After questions are pretested, each question is included in a statistical analysis to determine if there are any unfair group differences in the performance of individual test questions; if so, the question is removed from the pool of questions. Once questions pass this level of rigor, they are included in the test. Even then, statistical analyses will be repeated regularly to further ensure fairness.