NotesFAQContact Us
Collection
Advanced
Search Tips
ERIC Number: ED400294
Record Type: Non-Journal
Publication Date: 1991-May
Pages: 33
Abstractor: N/A
Reference Count: N/A
ISBN: N/A
ISSN: N/A
Issues of Test Bias, Item Bias, and Group Differences and What To Do While Waiting for the Answers.
Scheuneman, Janice Dowd; Slaughter, Carole
A number of explanations have been offered for the differences in test performance among various population subgroups. This paper begins with a discussion of these explanations including the psychometric explanation that group differences are due to bias in the test. An overview of bias research argues that results to date are inconclusive. A theory of bias is introduced that provides a definition of bias and a framework that explains why the issues are so difficult to resolve. Bias is defined as the systematic over- or underestimation of the true abilities of a group of examinees formed according to some demographic variable such as sex or ethnicity. The framework also provides a connection between test bias and item bias. The concept of item bias is then distinguished from that of differential item functioning (DIF). DIF research is described as promising in many regards, but with major areas of uncertainty in the interpretation of its results. It has generally supported the reliability and validity of standardized tests for minority groups, but as long as research is based on hypothetical scenarios instead of solid, research-based theory, the question of whether test bias accounts for some portion of the observed differences between groups is likely to remain unanswered. Practical guidance is offered for those who must make important decisions about individuals without knowing the answers about test bias. (Contains 26 references.) (Author/SLD)
Publication Type: Reports - Evaluative
Education Level: N/A
Audience: N/A
Language: English
Sponsor: N/A
Authoring Institution: N/A