Erin, J. N., Hong, S., Schoch, C., & Kuo, Y. (2006). Relationships among testing medium, test performance, and testing time of high school students who are visually impaired . Journal of Visual Impairment & Blindness , 100 (9), 523–532. https://doi.org/10.1177/0145482X0610000904

Also downloadable from ERIC online database: https://eric.ed.gov/?id=EJ746002

Oral delivery using an audio recording was compared with written forms of the social studies test. The written forms were assigned based on each participant's typical reading medium: standard font, enlarged font, and braille.

Twenty-seven high school students from three different U.S. states—Arizona, Iowa, and Tennessee—comprised this convenience sample. Nine students were blind, nine had visual impairments (low vision), and nine were sighted (had no vision-related or other disabilities). No additional demographic or other characteristic data were reported about the participants.

Scores were analyzed from six social studies classroom tests on the contents (six chapters) of a grade 8 social studies textbook. Each test consisted of 20 multiple choice and 20 short-answer items, and all of the social studies tests were developed by one of the study authors. Groupwise comparisons of performance means and group means for test completion time were analyzed. Student participants were interviewed, answering seven questions addressing their preferences about oral or written form and detailing their reasons, their perceptions of relative benefits of the forms, and relevant academic experiences.

Each of the three groups—students with blindness, students with low vision, and students with no vision-related disabilities—had no significant mean score differences between the oral delivery and written delivery forms of the full (40-item) social studies tests. Additionally, comparisons of total scores across the three groups detected no significant differences in performance by group membership. Further, these same patterns were found when comparing performance on short answer test items, both between form of delivery and by group membership. However, blind students scored significantly higher on the multiple-choice items when using braille than did the low-vision students when using enlarged font and than did the sighted students when using standard print. In contrast, all three groups performed similarly as one another on the short-answer items. Finally, all three participant groups scored significantly better on the multiple choice items than each group scored on the short answer items. Testing time did not vary by participant groups or by test forms alone; however, blind students took significantly longer to complete braille tests than tests delivered orally. Most test-takers thought that multiple-choice were easier than short-answer items. Nearly all test-takers preferred completing the written tests, indicating that doing so permitted them to engage directly with the material at their own rate. Only one student with blindness and two students with low vision preferred oral delivery, noting that doing so was easier and quicker; one low-vision student had no preference. When asked about perceptions of relative test speed, oral testing was identified correctly as faster by six blind students, two low-vision students, and one sighted student. In contrast, one blind student and one sighted student incorrectly perceived that they were faster with the written (braille and standard print) versions. Further, nearly all test-takers accurately identified the faster medium for each of them; only two sighted students and one blind student were incorrect in their perceptions. When asked about perceptions of score differences by test form, individual students' perceptions were frequently inaccurate; five sighted, three low-vision, and four blind students correctly identified their higher-scoring test forms. Further, slightly more students (n=14) scored better on the form they preferred, while 12 students scored better on their non-preferred test form (one student had no stated preference). These proportions varied only somewhat by participant group; however, most students with low vision thought that they performed worse on the oral delivery test form, yet they performed similarly across test forms. Limitations of the study were reported, and future research directions were suggested.