Accommodations for English Language Learners: Just How Effective Are They?

In all of our schools we face the issue of trying to determine how much progress English Language Learners (ELLs) are making in mainstream subjects, where their lack of language does not allow them to fully demonstrate their understanding of the concepts we want to assess.

We sometimes attempt to minimize the role of language in assessments by making accommodations for ELL students. These accommodations are often designed to remove or alter language in assessment tasks, which is not central to the construct being assessed. But do they really work? And are they valid?

A meta-analysis of testing accommodations for ELL learners was conducted recently, focusing on exactly these two questions. The analysis investigated the effectiveness, defined as “the extent to which students receiving the accommodation demonstrate improved test scores” (p. 1171), and the validity, defined as accommodations which do not affect the performance of students who do not need them, of a range of accommodations.

Only studies conducted in relation to large-scale testing of mathematics and science were included in the analysis. The studies included in the analysis involved 23,999 participants and included 38 separate tests of effectiveness.

The following types of accommodation were investigated:

1. Simplified English, where the language was changed to remove irrelevant linguistic complexity.

2. English dictionary or glossary, including provision of standard dictionaries, dictionaries customized to the test, or glossaries for specific words used in the assessment.

3. Spanish language test.

4. Dual language test, where the test was provided both in English and the student’s native language.

5. Bilingual dictionary or glossary.

6. Extra time—this was sometimes coupled with one of the other accommodations.

What Were the Results of the Analysis?

• The only accommodation found to have a statistically significant positive effect on the performance of ELLs was the use of an English dictionary or glossary. The effect was small, but did reduce the gap in performance between ELLs and non-ELLs by 24 percent in science and 20 percent in mathematics.

• All other accommodations were found to have no reliable effect on performance.

• There was a lot of variability in the results of studies using Spanish language tests, bilingual dictionaries and dual language assessments. The researchers attribute this largely to the language of instruction. If the language of instruction was English, then these accommodations sometimes even had a negative effect on performance.

• All of the accommodations were found to be valid in that they did not affect the performance of non-ELLs.

What Do These Results Mean for Our Classrooms?

The results reported here converge with previous reviews in demonstrating the promise of using English dictionaries and glossaries to support ELL Students during assessments. The results also suggest that students perform better on an assessment conducted in the language in which they receive content instruction.

Thus the use of translation into other languages or that of bilingual dictionaries may be unhelpful, or even potentially damaging, to performance. Even the single effective accommodation only slightly closed the performance gap.

In general, the potential of accommodations presented in this study seems less than positive. One hypothesis put forward by the researchers to explain the general ineffectiveness of accommodations is the suggestion that “the learning of content knowledge is largely mediated by language” (p. 1188). If this is the reason, then the performance gap between ELLs and non-ELLs may largely be due to the fact that ELL students lack the conceptual knowledge of the subject area, not merely the language skills needed to express it.

Accommodations will not help us get around this. Our task remains the much more difficult one of finding ways to explicitly teach the language of the subject area, even as we teach the concepts of the curriculum.

The study

Kieffer, M. J., Lesaux, N. K., Rivera, M. and Francis, D. J. (2009). “Accommodations for English Language Learners Taking Large-Scale Assessments: A Meta-Analysis on Effectiveness and Validity.” Review of Educational Research, 79 3, pp. 1168-1201.

---

Please fill out the form below if you would like to post a comment on this article:

Nickname

(this will appear with your comments)

Email

Comments

---

Comments

There are currently no comments posted. Please post one via the form above.

MORE FROM

GORDON ELDRIDGE: LESSONS IN LEARNING