No Child Left Behind (NCLB) requires states to measure student performance toward state targets both in the aggregate for "all students" as well as for every subgroup. One such subgroup is English language learners (ELLs). The AFT has long supported disaggregation of data. Without it, schools can appear to be doing a good job overall when, instead, they are inadequately serving particular subgroups of their students, for example their English language learner population. In addition, analysis of disaggregated data is an important process in helping schools know where more help is needed. 

The AFT believes that ELL students must be included in the NCLB accountability system, but in an appropriate manner. In February 2004, the U.S. Department of Education made two important policy changes with respect to the ELL subgroup and adequate yearly progress (AYP) calculations. Specifically, the Department will now allow states to exempt students who are new to this country and to the English language from taking the reading/language arts content assessment for one year. (These students would still be required to also take a mathematics content assessment, with appropriate accommodations. States may, but would not be required to, include results from the math, and, if given, the reading/language arts content assessment in AYP calculations. These students will continue to count for NCLB's required 95 percent participation rate.) As required under Title III, ELLs will continue to be tested for English language proficiency. The second recent change announced by the Department will allow states, for up to two years, to include in the ELL subgroup students who have attained English proficiency.

While these policy changes are a productive first step, more has to be done to ensure that states and schools are held accountable for the academic achievement of ELL students—more than 4.6 million of whom attend school in the United States. Even with the recent federal changes, the law's AYP formula continues to pose two unique challenges for the ELL subgroup.

Challenge 1: How ELL subgroups are defined works against attaining proficiency
The first major problem revolves around how this subgroup is defined. Once ELLs become proficient in English, they move out of the subgroup, which means that the ELL subgroup functions as a revolving door. The recent policy change—allowing students who have exited the ELL subgroup to count in ELL AYP calculation for two additional years—helps this situation, but falls short of being a full remedy. ELL students should be counted in this subgroup for at least three years after they exit this subgroup. Research has shown that it can take between four and seven years to attain full proficiency in standard academic English. Therefore, additional years of achievement data will allow a more accurate demonstration of proficiency and academic performance.

In addition, the policy change that gives new immigrant students a one-year exemption from taking reading/language arts content tests should be extended to include students who were born in the U.S., but who are non-English speakers. Just like new arrivals to the U.S., these students also struggle with English when they arrive at schools and are rightfully considered part of the ELL subgroup.

Challenge 2: Too few native language state assessments exist
A second, related problem is that most states do not offer native language or linguistically modified assessments of academic skills to students who haven't yet mastered English. By definition, ELL students are not proficient in English. Until they become proficient in English, they will struggle with any test that is written in English. Therefore, any test written in English that seeks to measure students' skills in reading and mathematics will be inaccurate. Both research and common sense dictate that non-native language assessments negates the validity of the assessments, demoralizes ELL students who time and again score poorly on the tests because they can't understand them, and risks pitting ELL subgroups against the rest of the school as the reason why the school has failed to make AYP. (Each and every subgroup in a school must make AYP proficiency targets, or the entire schools is determined to have not met AYP.)

What's worse, NCLB is vague on states' responsibilities for developing native language or linguistically modified tests. (Development of new assessments in languages other than English is not likely to come quickly without federal incentives or requirements because of state budget crunches and impending NCLB requirements that all students in grades 3-8 and grade 10 be tested, which for many states means the development of additional tests in multiple grades.) The federal government has provided no additional funds to states for this purpose, nor has it offered any guidance about the role of the state in translating or modifying assessments "to the extent practicable"—a phrase in the law that, at best, is ill-defined and, at worst, provides and easy "out" for states to neglect this important activity. And without native language or linguistically modified assessments, states now find themselves in a Catch-22 situation: If a students can't read an assessments because of a language barrier, then the assessment is not a valid measure of the student's academic proficiency. But students will still be taking these state assessments, even though they serve no valid educational purpose, because NCLB requires a 95 percent testing participation rate for each and every student subgroup.

Changes Needed to Appropriately Assess English Language Learners
In assessing the academic proficiency of English language learners, states and the federal government must make some changes to get things right:

• They must attend to the conflict between NCLB's requirements that all students be tested and the fact that few states have linguistically modified or native language assessments, forcing most ELLs students to take a test in a language in which they are not yet proficient.
• They must resolve the inconsistency between NCLB's requirement that assessments be "valid and reliable" and the clear fact that testing non-English speakers in English to assess their academic proficiency is an invalid practice.
• State, with assistance from the U.S. Department of Education, must resolve the confusion that exists between the assessment of English proficiency and the assessment of academic reading and math skills.
• The U.S. Department of Education must extend its recent policy changes to allow states to (1) provide the same time exemption to U.S.-born ELL students as that given to immigrant ELL students, and (2) count as ELLs those students who've exited from the ELL subgroup for at least three additional years.

It is educationally unsound to assess students with invalid measures and unfair to judge an entire school based on such technically flawed results. The U.S. Department of Education needs to enforce the provisions of the law to ensure that ELL students are properly assessed and that they are provided with the adequate time and resources to succeed academically.