Speech-to-Text Considerations

Kinds of Speech-to-Text Services

All speech-to-text services strive to quickly translate auditory information into text on a computer screen or other display device. This text provides access to verbal communication and auditory information for individuals who cannot hear.

There are three broad categories of speech-to-text services: CART, Text Interpreting (i.e., C-Print® and TypeWell®), and Automatic Speech Recognition systems (ASR). See the "Comparison of Speech-to-Text Services Chart" for details about each of these.

Differences and Similarities

From the consumer's viewpoint, there are primarily three factors to consider in choosing a speech-to-text service:

1. verbatim-ness = percentage of exact words spoken presented in text; word-for-word
2. accuracy of content = percentage of ideas expressed by speaker that are clearly and accurately shown in the text
3. quantity of words to read = this is related to verbatim-ness, as the higher the verbatim-ness, the more words of text

Legal Aspects of Speech-to-Text

The right to communication access is guaranteed by federal and state laws. These include the Americans with Disabilities Act (ADA), Section 504 of the Federal Rehabilitation Act, and the Individuals with Disabilities Education Act (IDEA).

This access can be provided by a sign language interpreter, or by a speech-to-text service. The U.S. Department of Justice's interpretive guidance that accompanies the Title II Regulation of ADA explicitly adds "notetakers" and "computer-aided transcription services" to the list of auxiliary aids and services for individuals who are deaf or hard of hearing.

Read more on these federal laws at:

• Section 504: www.section508.gov/index.cfm?FuseAction=Content&ID=15
• IDEA: idea.ed.gov/
• ADA: www.ada.gov/
• U.S. Department of Justice's interpretive guidance: www.ada.gov/taman2.html

Sign, Speech-to-Text, or Both?

If a student uses sign language, he may need to decide whether an interpreter or a speech-to-text service is best for him. Institutions rarely provide dual services for a single student.

When deciding which service is appropriate, consider the following:

• Does the student understand sign language well enough to comprehend the anticipated level of the class content?
• Will the class include extensive vocabulary or other items that the student may want to learn through permanent text rather than via fingerspelling or sign?
• Are the student's reading skills adequate for learning quickly via text?
• Is the student comfortable using detailed class notes to study (if transcripts are available)?
• Has the student used a speech-to-text service in the past? Did he learn well with this service? Did he have any issues or problems?
• If the student does not use her voice in class, is she willing to type questions/comments for the speech-to-text service provider to voice for her?

Remote Speech-to-Text

When the consumer (student), the speech-to-text service provider, and the speaker (instructor) are in different locations, remote speech-to-text services may be a viable option. Here are three situations that might be handled with remote services:

1. The student and instructor are in class, but no local service provider is available.
2. The instructor and the service provider are in class, but the student is attending from a remote site.
3. The instructor, the student and the service provider are all in separate locations. For example, the instructor is in the classroom; the student is viewing via video conferencing or the Internet; and the service provider is in a third location.

On October 25, 2007, PEPNet sponsored a TeleTraining titled, "An Inside Look: Providing Remote Captioning Accommodations." The full transcript, PowerPoint and handouts have been uploaded to the PEPNet website, and are available at www.pepnet.org/training/train071025/

Some of the issues addressed in this TeleTraining:

• Questions about remote captioning services;
• Using remote captioning accommodations when a live captionist is not available;
• Technology and equipment required to use remote captioning services;
• Getting information directly from different remote captioning agencies simultaneously.

Providing Hard-copy Texts of Classroom Lectures

If your institution provides students or others hard copies of class speech-to-text transcripts, it is important to know and understand the institution's intellectual property policies regarding the sharing of notes.

Some institutions give hard copies or digital files of the notes to the student receiving the service, others do not. Here are suggestions for why an institution might consider giving hard copies to students and others:

• Transcripts of lectures can be used as complete classroom notes, preserving the entire lecture and all students' comments for subsequent review by deaf and hard of hearing students taking the course.
• Instructors may welcome the transcripts as a way of tightening their lectures and reviewing their students' questions and comments.
• Transcripts can be of value in tutoring deaf and hard of hearing students, enabling tutors to organize tutoring sessions in close accord with course content.
• Interpreters can use the transcripts to improve their signing of course-specific words and expressions.