Can One Size Fit All? The Imperfect Assumptions of Parallel Achievement Tests for Bilingual Students

Abstract

With the advent of George W. Bush's education policy, emphasizing the frequent large-scale assessment of children in American public schools, it has become even more important than ever before to examine the fairness of the testing process and instruments being used to make decisions about children and their schools. When the children being assessed have limited English proficiency, one of the most common means of assessing them is the use of parallel assessments: standardized achievement tests, developed in the native language of the English language learners, that emulate the content of their English-language counterparts. This article focuses on some of the faulty assumptions that are made in the development of such tests for Spanish-speaking English language learners and argues that new measures need to be taken to assure that these tests reflect the best interests of the populations to whom they are administered.