Effects of Accommodations on High-Stakes Testing for Students with Reading Disabilities

Abstract

The interaction hypothesis proposes that valid test accommodations benefit only those with disabilities. To evaluate this hypothesis, Grade 3 students with word decoding difficulties identified with dyslexia and average decoders were randomly assigned to take the same version of the Texas reading accountability assessment under accommodated and standard administrations. The accommodated administration was given in 2 sessions with oral reading of proper nouns and comprehension stems. Only students with decoding problems benefited from the accommodations, showing a significant increase in average performance and a 7-fold increase in the odds of passing the test. These results supported the interaction hypothesis, showing that accommodations designed for a clearly defined academic disability can enhance performance on a high-stakes assessment.