Need or Problem:

Most universities and colleges follow the same basic philosophy toward academic accommodation for disabilities: the student's disability is viewed as the barrier to their education; therefore, we provide adaptations so that the student can benefit from the existing educational environment.

Many such adaptations are necessitated because the student's disability limits the way they interact with academic content (course materials, texts, etc.) and instruction. Thus, accommodations often consist of converting content and instruction from one form to another. For example, printed material is provided in Braille for a blind student; an in-class stenocaptioner transcribes the lecture for a student with hearing impairments; a peer notetaker takes notes for a student who is physically unable to; a student with dyslexia receives printed material as e-text which computer software reads aloud. Accommodations generally take place on a case-by-case basis, are often specialized, and can be costly. In addition, students needing materials in alternative forms sometimes receive it after their peers do because of the additional time needed for acquisition and conversion.

Solution:

The Proteus Learning Environment adopts a more progressive approach by building these kinds of accommodations into the existing educational infrastructure. This is accomplished through the thoughtful, proactive application of curriculum design, teaching strategies, and technology. Unprecedented advances in innovative technologies makes it possible to design courses, augment instructional techniques, and present materials flexibly in a variety of forms. With technology accommodations such as those mentioned above embedded into the infrastructure of a course, and with curriculum designed to take advantage of these capabilities, students with disabilities and learning differences can compete more fairly with their peers. The goal is to augment the learning environment so that many disabilities are largely transparent to the instructional process.

This approach is commonly referred to as Universal Design (UD), a term originating in the fields of architecture and product design. UD is the philosophy of designing products and environments to be usable by all people without the need for individualized adaptation or specialized design. Well-integrated UD features usually go unnoticed: automatic door openers, variable height counters, signage in multiple formats, and wheelchair curbcuts. Such design features are incorporated for people with disabilities, but are used -- and appreciated -- by everyone.

Likewise, Universal Design for Instruction (UDI) is inclusive; it addresses the needs of students with disabilities but also benefits all students. Simply put, think of UDI as "educational curbcuts". A not-insignificant side benefit of such a learning environment is that it also creates innovative instructional possibilities for faculty and novel learning opportunities for students.

Building on the principles of UDI, Proteus capitalizes on current and developing technologies to create an inclusive learning environment. While primarily conceived to remove learning barriers for students with disabilities, Proteus aims to enhance learning both in class and outside of class for all students. By offering alternative means for interacting with information, students can use the learning modalities that best meet their learning preferences or needs. Integral to this effort is the engagement of faculty to design curricula which exploits Proteus' capabilities.

Building on the principles of UDI, Proteus capitalizes on current and developing technologies to create an inclusive learning environment. While primarily conceived to remove learning barriers for students with disabilities, Proteus aims to enhance learning both in class and outside of class for all students. By offering alternative means for interacting with information, students can use the learning modalities that best meet their learning preferences or needs. Integral to this effort is the engagement of faculty to design curricula which exploits Proteus' capabilities.

Design:

An important insight was the recognition that nearly all academic accommodations involving content conversion either are, or could be, in digital electronic form, e.g., closed captioning, Braille, and e-text. Whiteboard output can be captured relatively easily, as can notetaking by peers (via either typing or handwriting recognition). Once speech recognition can cope with the nuances of an animated lecture and not just controlled dictation, lectures can be captured as text transcript as well as digital audio. In the interim, remote stenocaptioning provides real-time text transcript of both lecture and classroom discussion.

Instruction and content are captured and, if necessary, converted to additional alternative forms (see diagram). These various "tracks" of information are then synchronized so that one can, say, search for a specific place in the lecture according to the spoken text, go directly to that place in the media file, and pull up one's personal notes taken at the time.

The initial test of a Proteus-enhanced course is planned for spring quarter 2005. A remote captioning service transcribes the instructor's lecture in real-time; this transcript appears in class for reference by students. Afterwards, students will be able to access a media file from Coursework which plays video and audio along with a synchronized text transcript (think "closed captioning"). A full transcript of the lecture will also be available. Because the spoken lecture will be available in an alternative form as text, students will be able to search for a particular word or phrase and go to each point in the lecture where it was used. Future versions of Proteus will allow for additional tracks of instruction and content as well as synchronization with each student's personal notes.

The table below illustrates how students with diverse and unique learning needs benefit from using the same basic system. This is UDI in action--the same environment meets every user's needs without individualized adaptation. This approach also promises to reduce the cost and delivery time of instruction and content in alternative formats.

Examples of How Different Students Might Use the Proteus Learning Environment

Scenario	Student Needs	How Student Might Use Proteus
Student #1: Learning disabilities, ADHD	Difficulty focusing on lecture while taking notes	Listens to lecture; Notes taken by peer notetaker (synchronized); Bookmarks notes track for later referral (e.g., Important, Don't understand, Ask professor, Assignment)
Student #2: Hearing impairment	Needs stenocaptioner, notes of lecture for later	Transcript of lecture via remote captioner (in future, via speech recognition); Annotates transcript in class (synched); Lecture transcript, closed captioned video, and notes available for review after class
Student #3: Legally blind	Does not have access to text and graphics used in class	Listens to lecture and bookmarks points; Can take notes on Braille notetaker (synched); Can use refreshable Braille display or text-to-speech software to read material from Web and alternative format material prepared previously
Student #4: Any student (studying after class)	Prefers to interact with information according to chosen learning styles	Configures text display to preferences (background/text colors, masking, etc.); Reviews bookmarked sections of stream for follow-up; Reads and listens to lecture transcript with text-to-speech screen reading software; Replays desired sections of video, whiteboard capture (synched with text for easy search); Annotates transcript and class notes using speech recognition, typing, or natural handwriting recognition

Challenges and Changes:

The biggest hurdle has been finding a realistic approach to convert classroom speech into text. Because so much instruction relies on the spoken word (whether lecture or discussion), it was imperative that we be able to represent this in a more accessible form. While current speech recognition technology is sufficiently advanced to accurately transcribe controlled dictation, it cannot deal with the vocal nuances and variables of an animated lecture, and is not yet capable of handling multiple speakers.

Our solution was to employ an accommodation frequently used for people with hearing impairments: remote transcription. Lecture audio is transmitted via voice-over-IP to a captioning service. Using a court stenographer machine, a specially trained captionist produces a transcript at up to 200 words per minute which is transmitted back and displayed real-time in class. A cleaned-up version of the text (available hours later from the service) is synchronized with the video and audio files.

Conclusions:

An important part of the testing phase is to evaluate how it impacts student learning, but also to examine how it augments an instructor's approach to their course--for example, "With these capabilities, how might I make different/better use of class time? How might I teach this material differently?" We are currently looking into funding opportunities to answer these questions, as well as further developing the underlying technology.