|Title:||Expanding Audio Access to Mathematics Expressions by Students with Visual Impairments via MathML|
|Principal Investigator:||Frankel, Lois||Awardee:||Educational Testing Service|
|Program:||Mathematics and Science Education: Special Education Research [Program Details]|
|Award Period:||07/01/2011 – 06/30/2014||Award Amount:||$1,498,052|
Purpose: Mathematical expressions are used in instructional materials, test-preparation materials, and educational assessments. These expressions pose an accessibility challenge for students with visual impairments because the information is difficult to convey using the available technologies, such as recorded or human-read audio. Existing assistive technology that provides synthetic speech for electronic text does not improve the accessibility situation for math because the technology generally does not "know" how to describe mathematical expressions.
The goal of this project is to enable math expressions to be presented usefully through assistive technology so that teachers and others can provide students with visual impairments timely access to classroom materials, tests, and test-preparation materials in a format that can improve their comprehension. The researchers will develop ClearSpeak, an accessible mathematical markup language (MathML) that can be integrated with existing screen reader software currently being used by the visually impaired community.
Project Activities: The researchers will develop and refine four components as part of the full intervention program. These components include: (1) standardized synthesized speech for rendering mathematical content (ClearSpeak); (2) navigation tools for students; (3) ClearSpeak integration capability with Microsoft Word; and (4) customizable authoring tools for teachers. Development and feedback cycles on the different components will overlap in time as the researchers develop, refine, and finalize each of the components. Data collection activities include focus groups with students and teachers, student math performance measures, and expert review of accessibility features related to mathematics content, Braille, and audio output.
Products: The products to be developed include the ClearSpeak language fully integrated into screen reader software (MathPlayer and MathType), publications, and presentations on study progress.
Setting: The research will take place in middle and high schools in New Jersey, Texas, Washington, and California.
Population: The participants in the study include approximately 75 students with visual impairments in grades 8 through 11 enrolled in Algebra courses as well as their teachers.
Intervention: The intervention to be developed focuses on ClearSpeak, a standardized synthetic speech rendering of math for instruction, test preparation, and testing. The ClearSpeak speech style will then be made compatible with current screen reader software such as MathPlayer and MathType. The intervention will also include navigation tools for students as well as authoring tools for teachers.
Research Design and Methods: The intervention will be developed through an iterative research and development process. The research will involve an initial development of the ClearSpeak speech style and implementation within existing software (MathPlayer). After the team develops a functional prototype, an iterative process of gathering feedback and refining the prototype will take place. The feedback studies will be designed to address accessibility, usability for student users, and usability for teachers to author their own accessible materials. Finally, a pilot study will examine whether there is initial evidence of the intervention's promise for improving mathematics access and outcomes by analyzing student gain scores on math assessments.
Control Condition: There is no control condition.
Key Measures: Data collection instruments include a cognitive interview protocol, a usability survey targeting both teachers and students, and two equivalent Algebra 1 assessments derived from released state assessments in mathematics developed by ETS.
Data Analytic Strategy: Pilot study student scores will be examined for direction and magnitude of change using analysis of variance. Data collected through teacher surveys on fidelity of implementation, the usability of the tool, and perceptions of student engagement will be transcribed and coded to inform revisions to the prototype. Researchers will also transcribe and code the types and frequency of teachers' requests for assistance as well as their responses to the structured interviews to determine what modifications to the tools, documentation, or training are needed to enable efficient teacher use.
Publications from this project:
Frankel, L, Brownstein, B, & Soiffer, N (2013). Navigable, Customizable TTS for Algebra. Journal on Technology & Persons with Disabilities Volume 1 . Retrieved from http://scholarworks.csun.edu/handle/10211.3/121942.