What are testing accommodations? Extra time, fairness, and what a score still means
Test accommodations are changes to how a test is given, such as extra time or large print, that remove a barrier without changing what the test measures.
Dr. Russell T. WarneChief Scientist
Share
Testing accommodations are changes to how a test is presented, answered, timed, or located that remove a barrier unrelated to the skill being measured, so that a person with a disability can show what they actually know. Large print for a student with low vision and extra time for a student with a reading disability are typical examples. The defining condition is that the score must still mean the same thing it means for everyone else.
That condition is where most of the controversy lives. This article explains the line between an accommodation and a modification, the four common categories, what the research on extended time actually found, why clinical IQ tests are almost always given under standard conditions, and how the law handles accommodations on high-stakes admissions exams.
Accommodation vs. modification: the construct question
Every test is built to measure a "construct," the specific ability or body of knowledge the score is meant to represent. The Standards for Educational and Psychological Testing, published jointly by the American Educational Research Association, the American Psychological Association, and the National Council on Measurement in Education, draw the line there. An accommodation is a change that leaves the construct, and therefore the comparability of scores, intact. A modification is a change that alters what is being measured, and a score obtained with one cannot be read against the ordinary norms.
The National Center on Educational Outcomes puts the same idea in plain terms: an accommodation lets a student participate "in ways that allow abilities to be assessed rather than disabilities," while a modification "alters what the test is designed to measure or the comparability of scores."
The same change can fall on either side depending on the test. Reading the questions aloud is an accommodation on a mathematics test, where decoding print is not the point. On a reading comprehension test, it removes the very skill being measured, so it becomes a modification. This is why accommodation decisions have to start with the question of what a test is for, which is the territory of construct validity.
The four common categories of test accommodations
State assessment programs in the United States usually sort accommodations into four groups, the scheme used in the Council of Chief State School Officers accommodations manual and in later National Center on Educational Outcomes guidance:
• Presentation: Changes in how the test reaches the student, such as large print, Braille, a screen reader, sign-language interpretation of directions, or magnification.
• Response: Changes in how the student gives an answer, such as dictating to a scribe, marking answers in the test booklet instead of a bubble sheet, or using a word processor.
• Setting: Changes in where the test happens, such as a separate room, small-group administration, special lighting, or a seat away from distractions.
• Timing and scheduling: Changes in how much time is allowed or how it is organized, such as extended time, extra breaks, or splitting a test across several days.
For English learners, accommodations such as bilingual glossaries aim at a different barrier. A meta-analysis by Kieffer and colleagues (2009) found that none of the seven accommodations studied threatened validity, but only English dictionaries or glossaries produced a statistically significant gain, and it was small.
What the research says about extended time
Extended time is the most common accommodation and the most argued over. The core question is whether extra time removes a disability-related barrier or simply gives anyone who receives it a higher score.
Researchers framed this as the "interaction hypothesis": a valid accommodation should raise scores for students with disabilities while doing little or nothing for students without them. Lynn and Douglas Fuchs proposed a looser test they called "differential boost," which requires only that students with disabilities gain more from the accommodation than their peers do.
The most cited review is Sireci, Scarpati, and Li (2005) in the Review of Educational Research. They found the literature too varied to support many general conclusions, because accommodations, implementations, and student groups differed so much from study to study. Two findings did hold up. Extended time "tended to improve the performance of all students," although students with disabilities tended to show relatively greater gains, and read-aloud accommodations on math tests helped some students with disabilities. In other words, the strict interaction hypothesis largely failed for extended time, while the weaker differential-boost version found partial support.
Not every study agrees even with that. Lewandowski, Lovett, and Rogers (2008) gave a timed reading comprehension test to high school students with and without reading disabilities under several time conditions. The students without disabilities benefited more from the extra time. Extended time did allow the students with reading disabilities to attempt as many questions as their peers had attempted under standard time, which is a narrower but still meaningful kind of fairness. Lovett's (2010) review of the broader evidence concluded that the way extended time is currently granted raises real concerns, while pointing toward better practices such as matching the accommodation to documented functional limits.
Why clinical IQ tests are given under standard conditions
Clinical intelligence tests such as the Wechsler scales are a special case. Their scores are interpreted against norms collected from thousands of people who all took the test under the same scripted instructions and time limits. A score means "this is how you performed compared with people your age who took the test exactly this way." Change the conditions mid-administration and that comparison no longer holds.
Federal rules reflect this. The IDEA evaluation regulation requires that assessments be "administered in accordance with any instructions provided by the producer," and the Section 504 regulation requires tests to be given "in conformance with the instructions provided by their producer." Speed is also part of what some subtests measure. Several subtests are timed, and the subtests behind the processing speed index exist to measure how quickly a person works, so extra time on them would change the construct. That trade-off is discussed further in our piece on timed vs. untimed IQ tests.
What examiners do instead is choose and interpret. The same regulations require that a test given to someone with impaired sensory, manual, or speaking skills reflect their ability "rather than reflecting the student's impaired sensory, manual, or speaking skills." In practice that means selecting an instrument suited to the person in advance (a nonverbal battery for a child who is deaf, for example), dropping or substituting subtests the manual allows to be replaced, and documenting any departure from standard procedure so the report flags which scores can be compared with the norms and which cannot. A report that says plainly "this subtest was not interpretable" is more useful than a full score table built on a compromised administration.
Accommodations on high-stakes exams: the ADA, Section 504, and flagging
Admissions and licensing exams are governed by the Americans with Disabilities Act. Under 42 U.S.C. 12189 and the Department of Justice regulation at 28 CFR 36.309, a private testing entity must give its exams so that results reflect "the individual's aptitude or achievement level" rather than a disability, must limit documentation requests to what is reasonable, and must give "considerable weight" to accommodations a person previously received, including those in an IEP or a Section 504 plan.
The history of "flagging," marking score reports as obtained under nonstandard conditions, shows how the law and psychometrics collided. After the Breimhorst v. ETS litigation, a Blue Ribbon Panel that included Stephen Sireci reported in 2002 that its majority favored ending the flagging of SAT scores earned with extended time, citing scientific, psychometric, and social evidence. The College Board announced that it would stop flagging beginning in the 2003-04 school year.
The Law School Admission Council held out longer. In a 2014 consent decree with the Department of Justice and California's Department of Fair Employment and Housing, it agreed to pay $7.73 million in damages and penalties, to permanently stop annotating the score reports of candidates who received extended time, and to grant, without further documentation, accommodations a candidate had already been approved for on other standardized exams.
Frequently asked questions
What is the difference between an accommodation and a modification?
An accommodation changes how a test is given without changing what it measures, so the score can be compared with standard norms. A modification changes the skill being measured, such as reading a reading test aloud, so the score can no longer be interpreted the usual way.
Does extended time give students an unfair advantage?
Research shows that extended time tends to raise scores for most students, with somewhat larger gains for students with disabilities on average. Whether that counts as unfair depends on whether speed is part of what the test is meant to measure.
Can you get accommodations on an IQ test?
Usually not in the sense of changing the rules mid-test, because the norms assume standard administration. Instead, a qualified examiner selects a suitable test in advance, substitutes subtests where the manual permits, and notes any departure from standard procedure in the report.
Do the SAT and LSAT still flag accommodated scores?
No. The College Board stopped flagging extended-time scores beginning in the 2003-04 school year, and the Law School Admission Council permanently ended the practice under a 2014 consent decree.
Do I need an IEP or 504 plan to get exam accommodations?
Not necessarily, but it helps. Federal regulations require testing companies covered by the ADA to give considerable weight to accommodations listed in an IEP or Section 504 plan.
The takeaway
A testing accommodation is legitimate when it removes an obstacle that has nothing to do with the skill being measured, and suspect when it quietly changes that skill. Extended time illustrates the difficulty. It raises scores for nearly everyone and for many students with disabilities somewhat more, so its fairness turns on whether speed belongs to the construct. Clinical IQ tests sidestep the problem by holding administration constant and moving the flexibility into test selection and interpretation. If you want a score that was produced under consistent, standardized conditions, you can take a full-length online IQ test designed around exactly that principle.
References
1. American Educational Research Association, American Psychological Association, & National Council on Measurement in Education. (2014). Standards for educational and psychological testing. American Educational Research Association. testingstandards.net
2. Christensen, L. L., Thurlow, M. L., & Wang, T. (2009). Improving accommodations outcomes: Monitoring instructional and assessment accommodations for students with disabilities. National Center on Educational Outcomes, University of Minnesota. nceo.umn.edu
3. Kieffer, M. J., Lesaux, N. K., Rivera, M., & 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), 1168-1201. doi.org
4. Fuchs, L. S., & Fuchs, D. (2001). Helping teachers formulate sound test accommodation decisions for students with learning disabilities. Learning Disabilities Research & Practice, 16(3), 174-181. doi.org
5. Sireci, S. G., Scarpati, S. E., & Li, S. (2005). Test accommodations for students with disabilities: An analysis of the interaction hypothesis. Review of Educational Research, 75(4), 457-490. doi.org
6. Lewandowski, L. J., Lovett, B. J., & Rogers, C. L. (2008). Extended time as a testing accommodation for students with reading disabilities: Does a rising tide lift all ships? Journal of Psychoeducational Assessment, 26(4), 315-324. doi.org
7. Lovett, B. J. (2010). Extended time testing accommodations for students with disabilities: Answers to five fundamental questions. Review of Educational Research, 80(4), 611-638. doi.org
8. U.S. Department of Education. (2006). Evaluation procedures, 34 CFR 300.304. law.cornell.edu
9. U.S. Department of Education. Evaluation and placement, 34 CFR 104.35. law.cornell.edu
10. Americans with Disabilities Act, Examinations and courses, 42 U.S.C. 12189. law.cornell.edu
11. U.S. Department of Justice. (2010). Examinations and courses, 28 CFR 36.309. law.cornell.edu
12. Gregg, N., Mather, N., Shaywitz, S., & Sireci, S. (2002). The flagging test scores of individuals with disabilities who are granted the accommodation of extended time: A report of the majority opinion of the Blue Ribbon Panel on Flagging. Disability Rights Advocates. dralegal.org
13. Fine, L. (2002, August 7). Test companies lower 'flags' on college-entrance exams. Education Week. edweek.org
14. Department of Fair Employment and Housing and United States v. Law School Admission Council, Inc., Consent Decree, No. 3:12-cv-01830 (N.D. Cal. 2014). archive.ada.gov
Hero image: a Perkins Brailler, by Joao GF, released CC0 (public domain dedication). Via Wikimedia Commons.
Take our professional IQ test
Want to know your IQ? Try the first ever professional online IQ test.
