Read-Aloud on a Grade-Level Math Assessment: Accommodation or Modification?
Ms. Ortega teaches a 4th-grade inclusive classroom of 26 students, 6 with IEPs. Devon, age 9, has an IEP goal for multi-step word problems. His most recent data: 92% accuracy on grade-level computation probes (three-minute, 20 items), 45% on mixed word problems without support, and 80% on the same problems when the teacher reads them aloud. He explains his reasoning accurately when asked orally, and his reading fluency is 58 words per minute correct on a 4th-grade passage (class median 112). The upcoming unit assessment is a 12-item constructed-response test that requires students to read multi-step problems, show work, and write a sentence explaining each answer. Ms. Ortega has 40 minutes of co-planning time weekly with the special education teacher, and Devon receives 30 minutes of resource room support four days a week. The principal wants all students to take the same assessment so scores can be compared across classes.
1. What is the most likely primary barrier for Devon on this assessment, and what is the defensible next step? Use the numbers to justify your answer.
Read the worked answer
The data isolate the barrier. Devon computes at 92% but drops to 45% on unsupported word problems and rises to 80% when the teacher reads the problems aloud. Reasoning is intact when he explains orally. The variable that moves the score is the reading load, not the mathematics, so decoding is the primary barrier.
Because the construct being measured is problem-solving, reading the problems aloud (or providing an audio version) is an accommodation: it removes a barrier unrelated to the target skill without changing what the test measures. Reducing the number of items, simplifying the numbers, or supplying answer choices would be a modification - it changes the construct and makes cross-class comparison meaningless, which is exactly the principal's stated concern.
The next step is to make the read-aloud/audio accommodation explicit and documented, deliver it consistently, and continue fluency-building in the resource room in parallel. Waiting for fluency to improve before providing access would deny him the curriculum in the meantime. Progress monitor with parallel 12-item probes, and keep a separate fluency measure to see whether the gap narrows.
2. How should Ms. Ortega and the special education teacher use the 40-minute weekly co-planning block in the two weeks before the assessment?
Read the worked answer
Use backward design from the test format, not general discussion. First, both teachers analyze the 12 items' reading demands - multi-step structures, comparative language ("how many more"), embedded numbers - and decide which supports are accommodations (read-aloud, extended time, separate setting) versus scaffolds that should be faded (graphic organizers, sentence starters). Second, build a bank of parallel problems so Devon can practice and be monitored in the exact format he will face. Third, write a read-aloud protocol so delivery is consistent: read each problem twice, no paraphrasing that simplifies the mathematics, no hints. Fourth, agree on the scoring rubric for work shown and written explanations, so the accommodation does not quietly loosen expectations. Fifth, coordinate with the resource room so the same item types and the same read-aloud routine are practiced there - an accommodation should never be novel on test day. Finally, plan the brief for the principal and the parent communication, and decide who monitors what. Paperwork can happen outside the block.
3. The principal objects that reading the test aloud makes Devon's score incomparable to other classes. How should the team respond?
Read the worked answer
Acknowledge that comparability is a legitimate technical concern, then distinguish the two mechanisms. An accommodation leaves the construct intact: it changes how the student accesses the items but not what counts as correct. Reading aloud removes decoding variance, which is not what a math problem-solving test claims to measure, so Devon's score remains interpretable as a measure of problem-solving - though not as a measure of reading. Standard practice is to report assessment results with accommodations flagged rather than to withhold access.
Offer concrete options. One: report Devon's score with the accommodation documented, and add a separate reading measure so no one over-interprets his math score as reading proficiency. Two: if the principal remains concerned, redesign the assessment for the whole class in two sections - a computation section and a problem-solving section delivered with audio or teacher read-aloud for everyone. Universal design removes the comparability objection entirely, because all students receive the same conditions, and it benefits other students with decoding weaknesses who have no IEP.