The objective of this article is to explore reports of unfair results for African American and other minority students on the SAT. Specifically, we studied the correlation between item difficulty and differential item functioning (DIF) values (i.e., African
Americans tend to perform better in harder items and White students perform better in easier items) reported by Kulick and Hu (1989) and Freedle (2003). Both investigations used the standardization approach to study DIF.1 In contrast, we will use item response theory (IRT) methods.