Learn About: Evaluating Performance | Common Core
Home > Instruction > What research says about English language learners > What research says about testing accommodations for ELLs
| print Print

What research says about testing accommodations for ELLs

Although our English Language Learners (ELL) research review focuses on instructional aspects of improving achievement for ELLs, assessment issues also play a role. Numerous studies have documented the fact that language and performance on achievement tests are confounded for ELLs more than for most students (e.g., Abedi, Lord, and Plummer 1997; Abedi 2003). States have included ELLs in large-scale achievement tests in various ways. At least eight states provide tests for some grades or subjects in languages other than English. Other states have used alternative assessments, such as portfolios, but after the U.S. Department of Education questioned whether the alternative assessments in reading and mathematics in eighteen states were comparable to regular tests, some states dropped the alternative assessments while other states began comparability studies. The most common approach, however, is to use regular state tests with accommodations (e.g., simplification of wording, provision of bilingual or English-language dictionary, extra time).

What does the research say about which accommodations are effective? Several studies have found linguistic modifications of test questions with excessive language demands to be effective (see Abedi and Dietel 2004 ). A recent meta-analysis from the Center on Instruction found that allowing students to use an English dictionary had the greatest effect, but the researchers noted that “for this accommodation to be successful in the testing situation, students must have experience with it during regular instruction.” The meta-analysis found a smaller average effect size when the accommodation added extra time. The researchers noted that the number of studies that met their criteria for inclusion was small (Francis, Rivera, Lesaux, Keiffer, and Rivera 2006c). Abedi (2003) found that translating test items from English to ELLs’ native language yields no significant increase in scores. He noted that testing students in a language other than the language of instruction may be the reason.

It is important to keep in mind that although these findings may be helpful in establishing policies, the needs and abilities of individual students must be taken into consideration in making decisions on a case-by-case basis. An accommodation that is appropriate for one student may not be appropriate for another.

This summary is based on a review conducted for the Center for Public Education by researchers at Edvantia, an education research and development not-for-profit corporation founded in 1966.  For more information, see Research Review: What research says about preparing English language learners for academic success.

Posted: October 30, 2007

©2007 Center for Public Education
Home > Instruction > What research says about English language learners > What research says about testing accommodations for ELLs