2000 Conference Proceedings

Go to previous article 
Go to next article 
Return to 2000 Table of Contents


Darlette S. Navrotski
Prentke Romich Company
Pittsburgh, Pennsylvania

Bruce R. Baker
University of Pittsburgh
Pittsburgh, Pennsylvania

Kenneth Kwasniewski
Pittsburgh Public Schools
Pittsburgh, Pennsylvania

Individuals with ALS confront many changes in their lives. Communication disability is one of the most significant problems. People in such a situation reasonably want to use a communication aid with the lightest learning cost and the greatest existing familiarity.

Various techniques for access have been developed to provide single switch scanning indefinitely for individuals with ALS throughout the course of the illness. Electronic augmentative communication systems often become slow and laborious as a person's physical abilities decline. Non-electronic eye-gaze systems often replace or supplement electronic systems.

The majority of electronic augmentative communication systems most often chosen for people with progressive neurodegenerative conditions incorporate language representation techniques based on letters. Because communication letter by letter can be tedious and fatiguing, rate enhancement techniques are used as well. Rate enhancement techniques designed for letters include abbreviation expansion and word prediction. Word prediction can further be divided into in-line prediction and columnar.

One's intuition that if an individual is literate, there is significantly less to teach, is fulfilled by the prescription of a word prediction system. The belief is that if a spelling system is prescribed, the individual can simply spell and do what is most familiar, and does not have to learn a different way of representing language. The ease of establishing a few common sense abbreviations conforms well to the typical clinical perception that initial ease of use is of great importance for people with neurodegenerative conditions. The intuitive claim of word prediction is very strong. After actuating only one, two, or three letters, one can then select the completely spelled word from an array by making only one more selection. Communication efficiency and rate enhancement were claimed as major benefits of keystroke reduction.

The popularity of abbreviation systems for broad based rate enhancement has declined since the late 80's. In those years, large packages of sentences abbreviated to suit a given user as well as abbreviations for common words were recommended. As word prediction became more popular, abbreviation expansion systems were combined with word prediction systems and used largely for prestored discourse cues and demographics. The bulk of the rate enhancement burden now generally falls to word prediction. Word prediction has been developed in augmentative communication for a long time. Beginning in the late 70's, word prediction systems, called, at that time linguistic prediction systems, were used to supplement Morse Code letter entry. By the late 1980's, a wide variety of word prediction systems for computers were available.

However, tight studies performed in the early 90's seem to question the rate enhancement component of word prediction systems. The works of Heidi Koester and Simon Levine at the University of Michigan clearly demonstrated that word prediction provided only marginal rate enhancement both for direct selectors and for scanners. Other researchers such as Horbail Venkatagari and others indicated that word prediction, while reducing keystrokes did not, in fact, increase communication rates significantly and in some cases resulted in actual rate decrements. The problem lay in the linguistic confusion resulting from the need to shift cognitive tasks from language generation to word recognition. Keystroke latencies grew to absorb the time saved by keystroke reduction.

Another rate enhancement technique, word-based semantic compaction systems, maintained rate enhancements equivalent to its keystroke reduction. These systems were generally not considered appropriate for people with neurodegenerative conditions because they depended on learning a new language representation technique (sequenced semantic icons with explicit grammar labels) and a significant investment in time.

With the failure of word prediction systems to provide significant rate enhancement, developers of semantic compaction word-based systems turned their attention to designing such a system for adult learning in a short time span.

The initial attempt (alpha test) involved using the Unity" vocabulary program and the DeltaTalker communication device, both manufactured by the Prentke Romich Company. Two individuals with the diagnosis of ALS and two individuals who acted as facilitators volunteered their time. The individuals with the diagnosis of ALS were both ambulatory, working, and their speech was intelligible to familiar listeners. The facilitators, both homemakers who had part time jobs, were instructed in the vocabulary program. Their training consisted of approximately 8 hours of hands-on time with an instructor. Both the individuals with ALS and the trainers were given their own device to practice with. The trainers were given no teaching materials/lesson plans and no direction for teaching the program. They were instructed to spend individual time with the person they were working with, and to teach the vocabulary and the use of the device. They were supported by someone who was familiar with the vocabulary and the equipment when questions arose.

Results revealed that both of the individuals with ALS learned the vocabulary in less than 12 hours of instruction. They both agreed, that as with learning any new language, or learning anything new, practice was the most important component of the learning. Both individuals were able to communicate using their DeltaTalkers, although, as with a beginning speaker of a second language, were not automatic with every attempt to recall a word. They were also still using speech as their primary means of communication. With practice, automaticity increased.

At the time of this writing, we are beginning beta testing. The Unity" vocabulary program was modified to include 44 "Quick Hits" - words that are retrieved by a single keystroke, or double hit on the same key. Other modifications in vocabulary and device functions were made to improve

efficiency. Lesson plans are being written with the goal that a facilitator can teach the vocabulary and device function in less than 12 hours with no training, or that the individual with ALS can learn the program independently.

This presentation provides the outlining of the learning approach as well as the proposed learning structures.


1. Blackstone, S. (ed.), Augmentative Communication News, Vol. 12, No. 1 & 2, 1998.

2 Vanderheiden, G. & Kelso, D. Comparative Analysis of Fixed-Vocabulary Communication Acceleration Techniques, Augmentative and Alternative Communication, pp. 196-206, 1987.

3. Horstman-Koester, H. & Levine, Learning and Performance of Able-Bodied Individuals Using Scanning Systems with and without Word Prediction, Assistive Technology, Vol. 6, pp. 42-53, 1994.

4. Higginbotham, D. Jeffrey. Evaluation of Keystroke Savings across Five Assistive Communication Technologies, Augmentative and Alternative Communication, Vol. 8., 1992.

5. Gardner-Bonneau, Daryl J. & Schwartz, Paul J. A Comparison of Words Strategy and Traditional Orthography, Proceedings RESNA 12th Annual Conference, New Orleans, Louisiana, 1989.

6. Forshew, D. & Hulihan, S. Living with ALS: Coping with change, The Amyotrophic Lateral Sclerosis Association, 1997.

Go to previous article 
Go to next article 
Return to 2000 Table of Contents 
Return to Table of Proceedings

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