Higgins, J., & Katz, M. (2013). A comparison of audio representations of mathematics content . Journal of Special Education Technology , 28 (3), 59–66. https://doi.org/10.1177/016264341302800305

Using the read-aloud accommodation presented through a computerized test format via text-to-speech software, the researchers compared the audio representation scripts presented in literal and interpretive forms.

Students (n=229) from 12 schools (3 in New Hampshire, 9 in Florida) in grades 5–8 and 9–12 participated; grades 5–8 participated in the study on tables, and grades 9–12 participated in the study on expressions or graphs with keys (students were permitted to participate in only one of the three studies). The three comparison groups included: (a) students with print disabilities or math access needs (n=124) such as dyscalculia who used oral delivery accommodations during assessments, (b) students with no access needs (n=102), and (c) students with low vision (n=5) who used large-print materials during instruction and assessments, and oral delivery accommodations during assessments.

Participant scores were compared on sets of released items from state mathematics assessments in Ohio, Virginia, Massachusetts, and Arizona. In each study, 10–​12 items were used in sets of two similar items each, in order to compare the audio representation of the items in both literal and interpretive forms. In addition to tracking the number correct when using each scripted form, student preferences regarding the forms were documented when students completed survey items immediately after the test items, and their opinions were also elicited by interviewers using cognitive labs, a semi-structured interview format. Due to the low number of students with low vision (n=5) in the three studies, data were only reported from the interviews and not the test scores.

Analysis of the performance data indicated that the students without disabilities varied as to which type of audio presentation (literal or interpretive) showed higher scores on average, across different types of items—those containing parentheses, exponents, tables, and graphs. Despite this variation, only one scoring difference was significant: students without disabilities scored significantly better when presented with the literal form for parentheses ("open/close parentheses"). In contrast, for exponents, students without disabilities scored only slightly better with the literal form; for tables, students without disabilities scored essentially the same for both forms; and for graphs with keys, students without disabilities scored slightly better with the interpretive form. Students with disabilities showed a similar pattern, scoring significantly better on parentheses items with the literal form of audio presentation. A different pattern emerges but not to a level of statistical significance, for the other types of items. On items with exponents, tables, and graphs, students with disabilities performed slightly better with the literal interpretation. Regarding student preferences about the audio script, the views of students without disabilities and students with disabilities were mostly similar, but sometimes different. During the surveys, majorities of all students (with and without disabilities) preferred the interpretive form for exponents and tables and the literal form for parentheses. However, a majority (69%) of students with disabilities preferred the literal form of the graphs with keys items. In contrast, about half of the students without disabilities preferred either the literal or interpretive form for graphs with keys. The interview data showed a few major themes. Study participants expressed comfort with using the computer platform and the embedded oral delivery software to access the assessment items. The test item content, at grade level, was challenging to participants. Some participants indicated that they were unaware, until it was pointed out, that there was anything different about the way the items were presented. Participants also indicated their thoughts and reasons for their preferences for specific literal or interpretive item forms. Limitations of the study were reported, and future research directions were suggested.