Evidence of Content Bias on the McCarthy Scales With Mexican American Children: Implications for Test Translation and Nonbiased Assessment

Content bias of the McCarthy Scales of Children's Abilities (MSCA) was investigated with 142 dominant English-speaking and 162 dominant Spanish-speaking Mexican American preschool children from a low socioeconomic status background. An item-group (partial correlation) method that controlled for age, sex, and ability was used to examine bias. Of the 157 MSCA items examined, 23 were found to be biased-the vast majority against the Spanish language group. The items biased against the Spanish-speaking children clustered in two subtests (Verbal Memory I and Numerical Memory I) that tapped serial-order short-term memory. This finding is discussed in the context of variations in information overload on memory as influenced by differences between the Spanish and English languages in word length and acoustic similarity. Implications concerning test translation, the issue of equivalence, and nonbiased assessment are discussed.