Differential Item Functioning and Item Bias: Critical Considerations in Test Fairness

Abstract

In recent years, policy makers, administrators, and test developers in the field of second language assessment have paid considerable attention to the issue of test fairness. A fair test is one that is comparably valid for all groups and individuals and that affords all examinees an equal opportunity to demonstrate the skills and knowledge which they have acquired and which are relevant to the test’s purpose (Roever, 2005). Various aspects of fairness in testing have been highlighted in the literature, including fairness in regards to standardization, test consequences/score use, and item bias (Kunnan, 2000; Shohamy, 2000). This commentary will focus, in part, upon the issue of item bias.