2002 Conference Proceedings

Go to previous article 
Go to next article 
Return to 2002 Table of Contents


VOICE RECOGNITION FOR STUDENTS WITH MILD DISABILITIES: GUIDELINES FOR TRAINING AND SUPPORT

Paula Walser, MS/CCC, SLP, ATP
Statewide Consultant
Wisconsin Assistive Technology Initiative
Cesa 6
PO Box 2568
Oshkosh, WI 54903
pwalser@cesa6.k12.wi.us 
920-236-0541

Talking to Write. How long have we eagerly waited for the technology to do what we wanted? How long have we waited for a product that school districts could afford? How long have we waited for portable computer technology to support voice recognition technologies? Wait no longer. Voice recognition technology is now affordable, requires operating systems available in most schools and households, and is a viable tool for students struggling with the mechanics of writing.

Students with mild disabilities often struggle with the mechanics of writing. These are the students who know what they want to say, but just are unable to demonstrate the coordination of putting the pen/pencil to paper or the fingers to the keyboard successfully. Voice recognition is one of many types of assistive technology that can assist and become the pen/pencil for these students.

What is voice recognition technology? It is a computer application that allows input via voice. Speech replaces the keyboard. Users can execute commands and write in conjunction with a word processing program. It works on both Mac and PC platforms. Now it also works well on both desktop and laptop models.

How does it work? User must dictate into microphone and teach the system to recognize his voice. As user speaks, software creates a user specific voice file which contains information about the users voice qualities and pronunciations. The system uses this information to make its best guess at what each word is as it is dictated. Training takes time to properly create a strong and accurate voice files. 100 % accuracy will never be achieved with current technology. Training time requirements have been greatly reduced with newer releases of the software. What used to take hours now takes virtually minutes.

Historically, we have had two types of voice recognition: discrete and continuous. Discrete requires the user to speak one- word- at- a- time. Continuous allows the user to speak in a somewhat continuous manner. Many of the newer versions of today's software allow the user to speak discretely within the continuous programs.

How fast can a person type using voice recognition software? This varies greatly from user to user. Most adults average 45-65 words per minute. This is also a difficult skill to measure since it does not take into account mistakes with the translation. Students with language disabilities may have reduced typing rates considering the processing skills that are involved.

Research: 1995 - Higgins & Zvi, Northridge University - LD college students using speech recognition software to complete written proficiency exam - with Speech Recognition scored same as peers without SR -with assistance of a transcriber scores fell below peers

Given the "absence of reflexive phonemic skills" (Campbell, R. & Butterworth B.1985), and that "at no point do they achieve appropriate age or reading levels of these skills" (Bruck, M. 1992), voice recognition offers a way in which students with learning disabilities can write without needing to do difficult phonemic processing. The smallest unit that needs to be processed is the whole word.

In the article," The Effects of Continuous vs. Discrete Speech Recognition Systems on the Reading and Spelling of Children with Learning Disabilities", Journal of Special Education Technology. 15 (1), Winter, 2000, by Eleanor Higgins and Marshall Raskind, The authors conducted a study with 39 students with Learning Disabilities between the ages of 9-18. They had one group of students use discrete voice recognition and another group of students use continuous voice recognition. Findings indicated that both discrete and continuous groups showed significant improvements in word recognition and reading comprehension. While the discrete group also showed significant improvements in spelling.

Who Can Benefit from this technology? Definitely visually impaired individuals, individuals with learning disabilities that interfere with ability to complete written language, children with physical disabilities restricting use of hands to keyboard or mouse, students with Traumatic Brain Injury, and etc.

An important consideration for selecting students for voice recognition is the fact that dictation is a new and different skill than writing. Many of the students we are attempting to use voice recognition with have never been successful with written language. They may have in fact never received instruction in written language that they were able to process and use. In some cases, we will also need to teach composition of written language, dictation skills, and how to use the voice recognition software.

Candidates for Successful use of Voice Recognition Software:

Within the state of Wisconsin, we have provided regional full day workshops on the topic of voice recognition. These workshops have been well attended by parents, students, teachers, and therapists of all kinds. Individual who participate in the workshop sit through a lecture about voice recognition. Then the participants each train their own voice's using the software. This is followed by a tutorial on the features of the software. Then we discuss suggestions for introducing voice recognition to their students. A majority of the teachers who attend the workshop leave with different students in mind for use of the technology. They also report feeling like they have a plan for implementation and places to go for support when using the technology.

Resources:

* "There's a Dragon in My School", http://www.the-literacy-center.com/chapterexrpts.htm. Case Study Excerpts From the Chapter, "There's a Dragon in My School" (c)1999 Shelley Lacey-Castelot, Dragon NaturallySpeaking(tm) Complete, Case Studies of elementary school children who have learning disabilities and who learn to use voice recognition software.

* http://www.keyspell.com - specializes in programs that enhance voice recognition programs

* http://www.InsyncIT.com - custom microphones and how to customize software

* http://www.norcom-electronics.com - transcription tools to use with voice recognition, portable recorders

* http://www.edc.org/spk2wrt/ - Funded by Department of Education, to explore the use of speech recognition software (or voice recognition) for writing by secondary students with disabilities.

* http://www.synapseadaptive.com/joel/default.htm/

* http://www.voicerecognition.com/

* http://www.naturalspeak.com/

* http://www.speechtechnology.com/

* http://www.out-loud.com/features.html


Go to previous article 
Go to next article 
Return to 2002 Table of Contents 
Return to Table of Proceedings


Reprinted with author(s) permission. Author(s) retain copyright.