Universal design and accommodations: accessibility, reliability, and validity

A recent trend in education research and policy has been to employ Universal Design principles for the greatest number of students to equitably and efficiently participate in large-scale assessments. A partnership was made with a computer science team to create a computer-based reading comprehension test with options for audio presentation and extended time testing accommodations. Tests with and without accommodations were given to 131 twelfth grade students in three high schools in New Jersey. Scores on Oral Reading Fluency Curriculum Based Measures (ORF CBMs) determined placement of students in groups of Students Without Functional Impairment in reading fluency (SWOFI, n = 83) and Students With Functional Impairment (SWFI, n = 44). Accessibility was measured using Test Accessibility and Modification Inventory (TAMI; Beddow, Elliot, & Kettler, 2008) and student evaluations, with TAMI results indicating an adequately accessible test and students rating the overall platform as highly accessible. Reliability change was analyzed using Feldt’s test, but yielded mixed results for different forms of the test. The validity results were mixed as well, with data analyses indicating no increase in scores for either group with accommodations, no differential boost, and no relationship between boost and accommodations. SWFIs spent more time and used more accommodations than SWOFIs, indicating that measurement of access skills can predict student need and use of accommodations. The relationship between reading fluency and reading comprehension lessened for SWFIs with accommodations, indicating that the accommodations may have assisted in removing construct-irrelevant skill. The results of this study suggest further inclusion of testing accommodations into general test platforms as an option for all students.Psy.D.Includes bibliographical referencesby Leah Dembitze