Ferrell, K. A., Correa-Torres, S. M., Howell, J. J., Pearson, R., Carver, W. M., Groll, A. S., Anthony, T. L., Matthews, D., Gould, B., O’Connell, T., Botsford, K. D., Dewald, H. P., Smyth, C. A., & Dewald, A. J. (2017). Audible image description as an accommodation in statewide assessments for students with visual and print disabilities . Journal of Visual Impairment & Blindness , 111 (4), 325–339. https://doi.org/10.1177/0145482X1711100403

Also located on organizational (American Foundation for the Blind) webpage https://www.afb.org/publications/jvib

The impact of three different accommodated assessment conditions was investigated: tactile graphics, audible descriptions of images, and a combination of tactile graphics and audible descriptions. This investigation was the work of an Enhanced Assessment Grant (EAG) project led by Utah, and involving departments of education in Colorado and Kansas, along with the National Center for Accessible Media (NCAM) and the National Center on Severe and Sensory Disabilities (NCSSD), examining "the use of audible image description as an assessment accommodation for students with visual and print disabilities" (p. 327).

Students with visual impairments (40% of all participants) and students with print disabilities (60% of all participants) in grades 3 through 8 from three states in the Western US (Colorado, Kansas, and Utah) participated; there was a total of 295 student participants. The group of students with visual impairments was tracked in terms of whether they could read standard-sized print, large or magnified print, and braille, and were provided test items in the appropriate format. Print disabilities included learning disabilities, autism, speech/language impairments, physical disabilities such as mobility-related disabilities, and other health impairments (these did not appear to include students with attention problems, because this specific difficulty was not identified). To be clear, student participants with the capacity to read print—including students with print disabilities and some students with visual impairments (but not complete blindness)—received test items in two conditions: with the audible image description accommodation and without that accommodation. [Note: Some of this participant group also received large print.] The students who read braille (9.5% of all participants) completed the test in three conditions: with the tactile graphics accommodation only, with the audible image description accommodation only, and with both tactile graphics and audible image description accommodations together.

The test items—comprising the academic content of English language arts (ELA), mathematics, and science—were drawn from previous non-secure and practice item banks in Utah. The ELA and math items were matched to the Common Core State Standards, and the selected science items were confirmed to link to grade level standards in all three states. The item response format was "forced response," apparently multiple choice, so that they would yield simple objective performance scores. Teachers also reported, through a brief survey, some basic information about each student participant: disability, assessment accommodations used, and current reading status—that is, teachers' indication of whether students are reading at grade level.  

Audible image description did not affect the performance of students with print disabilities—that is, this participant group's average score was not significantly different with this accommodations than without this accommodation. Students with visual impairments who read print—either standard-sized or large print—also did not score significantly differently when receiving audible image description than when they did not. However, students with visual impairments who read braille answered significantly more items correctly with audio description across all items, with a moderate effect size. When analyzing the score patterns for students who read braille more closely, audio description alone was the most beneficial of the three testing conditions, especially for ELA and math; braille readers answered correctly on the same proportion of science items when receiving tactile graphics only as when receiving audio description only. The researchers expected to find that audible image description would work together with tactile graphics to boost scores of students with visual impairments who used braille; however, this testing condition (with both accommodations) yielded lower scores than audio description alone in each of the three content areas for braille readers. In fact, in ELA and science, braille readers scored lowest in the multiple accommodations condition, compared with receiving tactile graphics only or audio description only. The researchers recommended that states ought to adopt audible image description as an accommodation for state assessments, in that it can be potentially helpful for students who read braille. Limitations of the study were reported, and future research directions were suggested.