Synthesizing the Effects of Test Accommodations for Special Education and Limited English Proficient Students

NICHCY’s Structured Abstract 20 describes the following:

Title | Synthesizing the Effects of Test Accommodations for Special Education and Limited English Proficient Students

Author | Chiu, C.W.T., & Pearson, P.D.

Source | Paper presented at the National Conference on Large Scale Assessment, June 13-16, Snowbird, UT. (ERIC Document Reproduction Service No. ED 433 362)

Year Published | 1999

Abstract
Test accommodations for special education (SP) and limited English proficient (LEP) students have attracted much attention recently, because proper accommodations promote inclusion and allow students to perform optimally. A meta-analysis of 30 research studies found empirical evidence supporting the position that, with appropriate accommodations, SP and LEP students can increase their scores on standardized achievement tests. Compared to conditions of no accommodation, students increased their scores by an average of 0.16 standard deviation. Relative to general education students, accommodated SP and LEP students demonstrated an average accommodation advantage of 0.10 standard deviation. Interpretations of these average effects require careful analyses because of the variety of accommodations, the specific status of the students, and the varying implementations of the accommodations. Providing additional time or unlimited time is the most frequently investigated accommodation. Other accommodations investigated were assistive devices, presentation formats, response formats, test settings, radical accommodations, and combinations of accommodations. Age did not seem to be a factor; elementary and postsecondary students benefited from accommodations. Narrative descriptions are given of the situations in which positive and negative effects of accommodation appear to emerge. An appendix lists and summarizes the studies analyzed. (Contains 63 references.) (ERIC: Author/SLD)

Background
Special education students (SE) and limited English proficiency students (LEP) constitute a large part of the student population in the United States. In this era of accountability SE and LEP students have to participate in state- and district- wide assessments and are entitled to have accommodations. Test accommodations promote inclusion and help students improve their performance. However, providing test accommodations is a subject that provokes a lot of debate because of the scarcity of research evidence to support the effects, impact, and validity of test accommodations. Policymakers and researchers are trying to determine what constitutes an effective accommodation, to what extent an accommodation can improve the scores of target students, and if accommodations should be provided to only target students or to all students. In this meta-analysis these questions are addressed, and the findings are synthesized.

Back to top

Research Questions

  1. What constitutes an effective accommodation?
  2. To what extent can an accommodation improve the scores of target students?
  3. Should accommodation be provided to only target students or to all students?
  4. What makes an accommodation effective or ineffective?

Research Design
Meta-Analysis*

 

  • Number of Studies Included | 30
  • Number of Subjects | N/A
  • Years Spanned | 1986-1998

Back to top

Research Subjects
The majority of student participants were in the K-6 level (n=8) and in the postsecondary level (n=16). Nine studies focused on middle and high school students, and four studies selected students from across a wide range of levels.

Age/Grade of Subjects
Age/grade ranged from kindergarten to postsecondary.

Specified Disability
Visual impairments, hearing impairments, physical disabilities, learning disabilities, hyperactive students, and students at risk

Intervention
Seven accommodations were investigated across the studies: timing of test, radical accommodation, presentation format, combination of accommodations, assistive device, setting of test, and response format. Three research designs were used:

  1. Repeated measure with a comparison group (n=16). The target population students were tested in both standard and accommodated conditions, and regular education students with similar backgrounds, grade levels, and IQ were selected as a comparison group. They were also tested in both conditions.
  2. Repeated measure without comparison (n=7).  This design was used for accommodations that were illogical to offer to non-targeted students.
  3. Equivalent group design (n=7). This approach consisted of studies in which students from the target population were randomly divided into two groups. One group was tested under the standard condition, and the other group was tested in the accommodated condition.

Duration of  Intervention
N/A

Back to top

Findings

  1. Timing of test was the most frequently investigated accommodation (47%). The LD subgroup was most widely studied (61%), followed by the subgroup of students with limited English proficiency (LEP) (16%).
  2. Overall, accommodations had a positive effect on the target population and an almost zero effect on general education students.
  3. Target students increased their scores by an average effect size of 0.16. General education students by comparison increased their scores by an average effect of 0.06.
  4. The results also showed that improperly implemented accommodation(s) could be disadvantageous to either the target population or the regular education population.
  5. Extended time or unlimited-time accommodation was beneficial to target populations when compared to standard time conditions.
  6. Evidence indicated that target populations might benefit from receiving a brief period with “practice items” to lower their anxiety level and to acquaint themselves with the test.

Combined Effects Size

  1. The overall effect size for all the target population students was 0.16, with a standard error of 0.02 (Q= 470.97, df =46, p< 0.01).
  2. For general education students, the effect size was 0.06, with a standard error of 0.02 (Q=392.83, df =30, p< 0.01).
  3. The relative accommodation effect (based on studies that provided empirical results for both populations) was 0.10, with a standard error 0.03 (Q=104.14, df=30, p< 0.01).

Interpretations of these average effects should be analyzed carefully because of the variety of accommodations, the specific status of the students, and the varying implementations of the accommodations.

Back to top

Conclusion/Recommendations

  1. The accommodation that has been most studied is extended time on tests and is, in general, the preferred accommodation.
  2. A remarkable amount of variation existed in the different accommodation categories.
  3. Sometimes target students benefited from accommodations, sometimes they did not, and sometimes they were disadvantaged by the accommodation.
  4. Findings also suggest that the idea of providing accommodations to “all students” to diminish the controversial issue of who should be accommodated is not appropriate, and could be harmful to either the target population or to regular education students.
  5. Standard conditions and the implementation of the accommodation are 2 factors that might form the basis of some guidelines for determining a proper way to administer accommodations.

______________

* Meta-Analysis
A widely-used research method in which 1) a systematic and reproducible search strategy is used to find as many studies as possible that address a given topic; 2) clear criterion are presented for inclusion/exclusion of individual studies into a larger analysis; 3) results of included studies are statistically combined to determine an overall effect (effect size) of one variable on another.

Back to top