Published by the National Center on Educational Outcomes
Number 11 / May 2000
Any or all portions of this document may be reproduced and distributed without prior permission, provided the source is cited as:
Thurlow, M., & Wiener, D. (2000). Non-approved accommodations: Recommendations for use and reporting (Policy Directions No. 11). Minneapolis, MN: University of Minnesota, National Center on Educational Outcomes. Retrieved [today's date], from the World Wide Web: http://cehd.umn.edu/NCEO/OnlinePubs/Policy11.htm
In most states, students with disabilities who participate in state assessment systems are permitted to use accommodations during testing as determined by their Individualized Education Program (IEP). Accommodations enable the students to demonstrate their knowledge and skills more effectively by reducing the effects of the disability. The right to use accommodations is defined in federal laws and supported by state statutes.
The term "accommodation" when used for testing generally refers to a change in procedures or materials that does not change the construct being tested or the comparability of scores obtained from accommodated and non-accommodated testing. However, there are some changes in testing that may alter the construct being tested or the comparability of scores. A commonly cited example is reading aloud a reading test to a student when the purpose of the test is to measure decoding skills. In some states, these types of changes are referred to as modifications; in other states, they are referred to as non-allowed or non-approved accommodations, or as nonstandard administrations.
What happens when students use these non-approved accommodations? How should test results be used and counted? Should students be permitted to use any accommodations, even those that are used in instruction but are not approved by states for testing? Should these students be considered as having met the standards? Should their test scores be reported in the same way as other scores are reported?
This NCEO Policy Directions is written to assist states and districts as they consider the issues surrounding the use of non-approved accommodations. Recommendations are provided for ways to use and report data from assessments taken with non-approved accommodations.
There are many terms used to refer to changes in testing procedures or materials. "Accommodation" is a general term referring to changes in the setting in which a test is administered, the timing of a test, the scheduling of a test, the ways in which the test is presented, and the ways in which the student responds to the test. Almost universally, the term is used to refer to changes that do not alter in any significant way what the test measures or the comparability of scores. Changes in setting, timing, scheduling, presentation, or response that are considered to alter what the test is supposed to measure often are called "modifications," "nonstandard administrations," or "non-approved accommodations."
This differentiation among the terms is not universal. In some states, the terms "modification" and "modified" do not refer to non-allowable changes, and have nothing to do with the perceived appropriateness or acceptability of the changes. The critical issue for using and reporting data is not what the change is called, but rather, whether it is considered to alter the skill being measured or the ability to compare scores.
Determining which accommodations to allow (because they provide comparability) and which not to permit (because they change what is being tested) has been the subject of ongoing research and much debate. Not everyone agrees on what constitutes a change that either alters what is measured or the comparability of scores. Research has not provided conclusive answers on many of the accommodations in question either, since studies on several have proven to be unreliable or contradictory, while research on others has not yet occurred.
The assumption surrounding non-approved accommodations is that a score from an assessment taken with a non-approved accommodation is different in some way from a score from a standard or approved accommodated administration. This often means that the score from the non-approved accommodation will be excluded, or changed in some way, or will be considered in a separate scoring category, rather than factored into standard accountability measures.
There are two important considerations in the controversy surrounding non-approved accommodations. First, the determination of what accommodations should be allowed or non-allowed is open to interpretation. Second, states must decide what will be done with test results if a non-approved accommodation is used by a student, especially when proficiency is exhibited under these conditions.
Most states distinguish between "allowable" and "non-allowable" accommodations or between "standard" and "nonstandard" accommodations. Some states use the term "aggregatable" and "nonaggregatable" to make the same distinction. In most cases, the distinctions are made in reference to specific tests or subtests. For example, reading the test to the students may be an approved accommodation for the math test, but a non-approved accommodation for the reading test.
States policies are complicated further by the different purposes of the tests, as well as by the use of different terminology. Terminology used by states in 1999 to identify test changes are summarized in Table 1.
In general, it appears that policies about accommodations are less stringent for criterion-referenced tests compared to norm-referenced tests, and for graduation exams compared to accountability exams. States seem to be searching for ways to provide all necessary accommodations to students, especially in cases where high stakes are involved. This practice seems to be driven by federal laws about the right of individuals with disabilities to use accommodations that they need. However, questions persist about whether students are automatically entitled to all accommodations listed on their IEPs, or whether states have the authority to limit the use of certain accommodations that have been determined to alter the assessment.
While many states agree in their designation of certain accommodations as changing what is tested, the use of others is less clear-cut, and their use is permitted in some states and denied in others. The use of pre-drawn graphic organizers to assist students during a composition test, for example, or use of manipulatives for a math test, are not universally agreed-upon, and therefore make comparisons among states difficult. Looking to individual states to see how non-approved accommodations are used, scored, and reported is not very enlightening because most are in the process of formulating their policies.
The most common policies for dealing with scores from tests taken with non-approved accommodations are to: (1) exclude the scores completely, and not report them, or (2) exclude the scores from aggregate reporting and accountability measures, but report them to the student and school, or (3) report the scores in an entirely separate category, or merged with results of alternate assessments. Few states are yet establishing what may be a more reasonable policy—collecting additional evidence for students who use non-approved accommodations, and using the scores from the additional body of evidence in place of the "non-approved" scores.
Table 1. Terminology Used for Test Changes
| State | Terms Used | State | Terms Used |
Alabama |
Accommodation; Nonstandard Administration | Montana | Accommodation; Modified Assessment |
Alaska |
Accommodation | Nevada | Accommodation |
Arizona |
Accommodation | New Hampshire | Accommodation |
Arkansas |
Allowable Accommodation | New Jersey |
Accommodation/Modification |
California |
Accommodation, Standard/Nonstandard | New Mexico |
Modification |
Colorado |
Accommodation, Documented/ Nondocumented | New York |
Modification |
Connecticut |
Allowable Accommodation | North Carolina |
Accommodations |
Delaware |
Allowable Accommodation | North Dakota | Accommodation |
Florida |
Test modification; Administration modification | Ohio |
Adaptations; Modification |
Georgia |
Accommodation, Standard/Nonstandard | Oklahoma | Accommodation |
Hawaii |
Allowable Accommodation | Oregon | Allowable Accommodation; Modification |
Idaho |
Accommodation | Pennsylvania | Accommodation |
Illinois |
Accommodation | Rhode Island | Accommodation |
Indiana |
Accommodation | South Carolina |
Accommodation; Modification |
Kansas |
Accommodation; Modification | South Dakota | Accommodation |
Kentucky |
Accommodation; Modification | Tennessee |
Modification/Accommodation |
Louisiana |
Accommodation | Texas |
Allowable/Nonallowable Accommodation |
Maine |
Modification | Utah | Accommodation |
Maryland |
Accommodation | Vermont | Accommodation |
Massachusetts |
Accommodation |
Virginia | Standard/Nonstandard Accommodation |
Michigan |
Accommodation | Washington | Accommodation |
Minnesota |
Accommodation; Modification | West Virginia | Accommodation; Modification |
Mississippi |
Accommodation; Nonallowable Accommodation | Wisconsin | Accommodation |
Missouri |
Accommodation | Wyoming | Accommodation |
Accommodations are changes that some students with disabilities need in order to have access to the test. Students who are blind and have not learned to read Braille essentially are denied access to the assessment if the test is not read to them, regardless of whether the test’s content is mathematics, reading, or some other content area. The same argument can be made for students with significant reading disabilities, and other conditions as well. Denying access to the assessment because of the effects of a disability, especially when the assessment provides access to a benefit (such as a diploma), raises many concerns.
Similar concerns are evident in some state policies where students with unique and significant physical disabilities are allowed the use of a scribe for the writing test as well as other tests, and where any student with a disability may use assistive technology during testing. It can be argued that a student who does not have use of his or her arms, for example, and who routinely uses "talk-and-type" software to express ideas in writing, should be able to use this accommodation on open-response and composition portions of the test. Still, assistive technologies often make assumptions about spelling, grammar, and punctuation, all parts of the scoring rubric in most states. Human scribes also may make these kinds of assumptions about written language when responses are dictated, such as assuming a pause indicates a comma or period. In cases where these language conventions count toward the score, the student using either assistive technology or a scribe might be considered to have an advantage over other students. One option currently used in some states is to disable grammar, spelling, and punctuation features on assistive technology, or to require students to dictate their responses and spell each word, identify upper case letters, and describe all punctuation.
When deciding what to do about the use and reporting of non-approved accommodations, states must determine (1) whether and under what circumstances students are permitted to use non-allowed accommodations, (2) how tests taken with non-allowed accommodations will be counted for system accountability (school scores, etc.) and student account-ability (graduation requirements), and (3) how scores will be reported.
Use of non-approved accommodations. Deciding simply not to permit the use of non-allowed accommodations is not recommended since it is not consistent with the Individuals with Disabilities Education Act (IDEA), which places decisions about accommodations in the hands of the IEP committee. Recent legal cases support this view. For example, in one case where there had been a denial of Read Aloud procedures, a settlement allowed two recently blinded students to have the reading test read aloud so they could earn their high school diplomas. Another case, involving the use of word processing aids (spell checkers) is still being decided.
Using an alternate assessment instead of providing for non-allowed accommodations may produce unintended consequences that are difficult to handle. Relying on an alternate assessment for students working toward a high school diploma, for example, may increase the number of students in the alternate assessment to levels higher than is considered reasonable (i.e., not a "small percentage"). In addition, it may place an unreasonable burden on educators.
Counting scores from non-allowed accommodations. It is not appropriate simply to discard scores from tests taken with non-allowed accommodations. One unintended consequence of doing so is an increase in the identification of need for non-approved accommodations for many more students than appropriate. This is especially the case when educators receive rewards or sanctions based on their students’ test results and feel threatened by high stakes accountability systems. In these instances, students who are viewed as poor performers or among those who are difficult to teach are likely to be determined to need special accommodations simply on the basis of those perceptions.
This tendency has been demonstrated time and again, so it is critical that every student count—even those who use accommodations that are not approved. Some states simply keep these students in the denominator when counting participation rates and when calculating performance averages.
If the option of "zero scores" is not acceptable, there is still a need to be careful about how scores from non-approved accommodations are counted, particularly if the accommodation seems to influence the construct assessed. For example, if a student has a reading disability, and the student uses a read aloud accommodation, we do not gain an accurate assessment of the student’s reading skills. If we were to simply use the score the student earned with the accommodation, we would be letting the school "off the hook" for making sure the student learns decoding skills.
Another possible unintended consequence, particularly if the scores were used for student accountability purposes, would be a likely increase in referrals to special education, since in most places, receiving special education services would make a student eligible for special accommodations. This could enable many more students to receive diplomas with the advantage of special accommodations, whether or not they were actually needed. Thus, we end up with a need to count, in some way, scores from assessments taken with non-approved accommodations. How this is done necessarily also affects how scores are reported.
Reporting scores from non-approved accommodations. How scores from tests taken with non-approved accommodations are reported depends, in part, on how the scores are determined. It is important to be clear about what has been done, regardless of the reporting strategy used. This means knowing how many students used these accommodations.
Permit students to use non-allowed accommodations, and collect and use additional evidence to obtain scores. Students with disabilities have a right to receive needed accommodations. Yet, the current status of test development and methodology, along with public perceptions, makes it unlikely that scores obtained when non-allowed accommodations are used will be perceived as valid and comparable. Thus, an approach is needed that bolsters information from tests taken with non-approved accommodations. Developing additional resources for implementation—possibly including the collection of a portfolio of information, use of performance assessments, or a panel review of evidence—makes sense for providing information needed to make decisions about student performance.
Whatever procedures are used, they should comprise a comprehensive and convincing body of evidence (perhaps including class grades, data on classroom performance, samples of student work, extracurricular activities, awards and other recognition) that documents that state standards have been met. In any event, it is advisable to use an independent group to review the evidence to make the final determination of a score.
How to report scores is still an issue, even if an additional body of evidence is collected to bolster information from tests taken with non-approved accommodations. Since the number of such scores should be small, it is probably most reasonable simply to include the score derived from the body of evidence with all other scores. At the same time, however, it is important to note the number of such scores included.
There will continue to be a need to determine which students really need non-allowed accommodations. Legally, this determination is to be made by the student’s IEP team. Still, specific guidelines for determining whether students need these accommodations would be beneficial for IEP teams. It might be useful to have a special request process in place by which IEP teams can inform the state of students needing non-allowed accommodations, and the reason the students need them. This process would discourage unjustified claims for special accommodations and at the same time enable the state to clarify its policies for reporting and accountability.
Before making any policy decision, however, it would be wise to estimate how many students would comprise the group of students who need non-allowed accommodations. Too large a number would indicate a need for additional guidance or training for IEP teams, and may necessitate creation of additional scoring categories for those students.
Recommended Approach for Non-Approved Accommodations |
| Permit students to use non-approved accommodations. Record these accommodation requests at the state level, and track the extent of their use. Collect and use additional evidence to obtain scores. |
There are currently no accepted "correct" answers to questions about how to use and report scores from assessments taken with non-approved accommodations. The approaches of individual states do not provide effective guidance at this point because most have not yet formulated their policies. States are aware they must soon decide whether to allow use of these accommodations, and whether to include, exclude, or flag scores in some way. Most states, however, are still in a climate of confusing terminology, policy, and practice, while awaiting clarification of their legal obligation to provide non-approved accommodations.
Ideally, the need to address the issue of non-allowed accommodations should occur relatively infrequently in practice. If large numbers of students are in need of these accommodations, then the appropriateness of the test or existing accommodations policies are suspect.
There does seem to be justification for permitting students to use non-allowed accommodations. The justification is based in instructional practice and federal law. At the same time, however, test developers have not figured out how to report the results when these special types of accommodations are used, and states must decide how these will count in meeting high-stakes requirements. Thus, states are left on their own for now to answer questions about these special situations.
Providing Assessment Accommodations (NCEO Policy Directions 7). Elliott, J., Thurlow, M., Ysseldyke, J., & Erickson, R. (1997). Minneapolis, MN: University of Minnesota, National Center on Educational Outcomes.
Reporting Educational Results for Students with Disabilities (NCEO Policy Directions 8). Erickson, R., Ysseldyke, J., Thurlow, M., & Elliott, J. (1997). Minneapolis, MN: University of Minnesota, National Center on Educational Outcomes.
Requirements for the Participation of Students with Disabilities in MCAS (Massachusetts Comprehensive Assessment System)—A Guide for Educators and Parents: 2000 Update. (2000). Malden, MA: Massachusetts Department of Education.
Testing Students with Disabilities: Practical Strategies for Complying with District and State Requirements. Thurlow, M., Elliott, J., & Ysseldyke, J. (1998). Thousand Oakes, CA: Corwin Press.
State Participation and Accommodations Policies for Students with Disabilities: 1999 Update (Synthesis Report 33). Thurlow, M., House, A., Boys, C., Scott, D., & Ysseldyke, J. (2000). Minneapolis, MN: University of Minnesota, National Center on Educational Outcomes.
This report was prepared by Martha Thurlow (NCEO) and Daniel Wiener (Massachusetts Department of Education)