Augmentative and alternative communication (AAC) technology is recommended for individuals with significant communication disorders, who may have motor limitations and other challenges that impact quality of life (QOL). The goal of AAC treatment is to optimize communication. This article presents innovations to the primary, secondary, and tertiary AAC components that are considered by rehabilitation clinicians when matching people with technology (MPT). Language considerations are paramount to the MPT process, and innovations are discussed on how features may enhance language performance. AAC technology has made performance and outcome gains contributing to the QOL of people who cannot speak.
Augmentative and alternative communication (AAC) is a field of endeavor addressing the expressive communication needs of people with significant speech disability. AAC interventions range from unaided methods using no technology (gestures, signs) up to high technology voice output communication systems. Over a lifetime, an individual may be recommended several AAC systems with the goal, as identified by the American Speech-Language-Hearing Association, to optimize communication for the highest quality of life (QOL) possible. Use of AAC technology has been shown to enhance QOL for children as young as 32 months, adults with amyotrophic lateral sclerosis (ALS), children with autism spectrum disorder, adolescents and adults with developmental disabilities, family networks, and adults with aphasia or other acquired neurologic conditions. The evidence base indicates that the life experience of people who use AAC is determined by their ability to achieve the highest performance communication possible. People who use AAC say that the two most important values in their use of AAC are saying exactly what they want to say and saying it as fast as they want.
Applying the principles of evidence-based practice (EBP) is expected to achieve the most effective communication. AAC stakeholders participating on AAC evaluation teams have become aware of the limitations of expert opinion as the basis for decisions on the selection of AAC technology. Yet, many individuals receive an unsystematic or idiosyncratic approach to AAC system selection. The model for matching the person with technology (MPT) systematically integrates personal, clinical, and external evidence into the AAC assessment process. By placing the client’s benefits first, practitioners pose specific questions of direct practical importance, gather objectively and efficiently the current best evidence, and take appropriate action guided by the evidence. EBP integrates personal data about an individual’s values, preferences, and expectations, with clinical data about the individuals abilities, skills, and performance, and an appraisal of the research evidence that matters. Consequently, AAC intervention is data-driven and consumer-centered. This poses a challenge to AAC rehabilitation, because practitioners must make clinical decisions that involve the identification, selection, and implementation of various AAC technology solutions based on evidence when minimal external evidence may be available.
Frameworks for improving quality of life
The World Health Organization’s International Classification of Functioning, Disability, and Health (ICF) provides a framework for representing the relationship among body or health status, activities, participation, and environmental and personal factors. AAC intervention can be placed within the ICF framework to show the interaction among the components influencing function. Identifying and defining the set of AAC system characteristics are critical to meeting the needs of individuals who rely on AAC and achieving long-term outcomes expected of the ICF framework. AAC system characteristics include a long list of technology features and components that are updated and released frequently that could influence activities and participation performance positively or negatively.
The MPT and the AAC language-based assessment models provide for a systematic and principled approach for selecting and evaluating AAC system characteristics in the face of continued progress in technology research and development. Each model reflects principles common to the ICF framework and EBP. The MPT is a set of person-centered measures that examine self-reported perspectives of individuals regarding strengths/capabilities, needs/goals, preferences and psychosocial characteristics, and expected technology benefits. In addition, a companion form is available so that provider perspectives can be assessed to ensure collaboration in the matching process. The AAC language-based assessment model conceptualizes a methodical approach to evaluate and select AAC interventions and technology. In a retrospective study involving an archival clinical data review, Hill and Scherer reported on the complex nature of the AAC assessment, including the MPT process, and the complex nature of selecting and integrating the full range of AAC system technology interventions. In addition, use of personal and clinical evidence provided data to support decisions when patient-oriented research evidence was lacking. Finally, the results showed that identified outcomes were achieved when AAC components were grouped based on patient priorities and influences on communication performance. The remainder of this article will address advances to primary, secondary, and tertiary components of state-of-the art AAC systems classified as QOL technology with the acknowledgment that not all innovations can be covered in the space provided.
Advances to primary components
The software designs of the AAC systems have made significant advances in performance because of improvements to how language is represented and generated using technology. The primary components of AAC technology are related to how the system can perform the functions of a natural language ( Fig. 1 ). Having the features of language available is critical to achieving the goal of AAC and providing the most effective communication possible. Language performance is influenced by the availability of the three AAC language representation methods (LRMs), the selection and organization of vocabulary, and the method of constructing messages/utterances. These language-based components of the technology generally influence communication performance more than any other components.
In evaluating the full range and large variety of AAC systems available, all AAC systems support only up to three language representation methods (LRMs): (1) alphabet-based methods, (2) single-meaning picture symbols, and (3) multiple-meaning symbols. The three AAC LRMs may be available singularly or as a combination of methods simultaneously depending on the software program and appearance of the visual overlay or display. Each of the three LRMs can be defined, characterized, and identified as available or unavailable on a system. Briefly, each is defined in the following sections.
Alphabet-based Methods
These use traditional orthography (letter-by-letter) and acceleration techniques that require literacy skills. The user selects from a standard keyboard, alphabet array, or whole word overlay to enter text to generate a message. Word prediction, while not an acceleration technique, is included as an alphabet-based method.
Single-meaning Pictures
These employ use of graphic symbols to represent one word or message. The symbols may be photos, line drawings, color graphics, or animated. A large symbol set is required to represent the typical spoken vocabulary, but literacy skills are not required. One subset of single-meaning pictures is visual scenes.
Multiple-meaning Icons
The use of a small set of symbols to select vocabulary and linguistic structures to generate messages has been termed Minspeak. Icons on a single page are used in a prescribed sequence to access a large vocabulary. Literacy skills are not required.
Vocabulary is selected and organized on AAC systems based on vocabulary frequency and language development and use. Vocabulary studies have confirmed that the high-frequency words used by speakers are consistent across cohorts and are a relatively small pool of words. These core words consist of about 450 to 500 words, but make up approximately 80% to 85% of the words used in a language sample regardless of the topic. Extended vocabulary consists of thousands of words that only make up 15% to 20% of the words used in a conversation related to a specific topic or activity. Modern vocabulary databases provide AAC software engineers with high-frequency vocabulary lists to support the selection and organization of core vocabulary. The most efficient AAC systems provide quick and random access to high-frequency vocabulary to avoid spelling and the need to search for locations. Also, single-display vocabulary organization allows word access to become automatic as in touch typing.
Methods of utterance generation are the third language component that influences the performance with an AAC system. Word-by-word construction of messages, when used on AAC technology, provides for spontaneous, novel utterance generation (SNUG). Spontaneous generation of self-created utterances is the hallmark of human language and corresponds to how children develop language and adults use language in everyday activities for full participation. SNUG allows individuals to say exactly what they want and intend to say. The design considerations regarding availability of the AAC LRMs and organization of core vocabulary influence SNUG. The use of prestored messages provides an alternative to spontaneous self-generated text and communication, but has several important performance issues.
Advances in computer speed and memory have provided the means to incorporate sophisticated algorithms for predicting language units longer than single words. Preprogrammed phrase and utterance-based programs may aid literate individuals who use AAC to engage in social conversations and structured situations. As the term implies, the quick-hit or quick-fire features, depending on the product, support pragmatic language functions. These one-hit prestored messages allow individuals convenient and prompt selection of prestored social comments, interjections, polite remarks, and conversational fillers to provide for turn-taking and topic maintenance during conversations. Fig. 2 shows an AAC system that has, among its other communication modes, programming to support social language use by prestored utterances and scripts appropriate for specific topics and situations or general conversational turn taking.
In summary, today’s AAC language software is multifaceted with complex programming that provides for the characteristics of a natural language. The MPT process starts with prioritization of the primary language considerations by ensuring the individuals are informed fully of all available options and performance information. Rehabilitation clinicians, however, must be careful when appraising the evidence base. For example, several studies have shown that word prediction is not any faster than spelling. Yet, word prediction is referenced routinely as a rate enhancement strategy in some product literature and textbooks. Because word prediction reduces the number of key strokes for text entry, the feature still may be recommended for patients who fatigue with AAC technology use. Fig. 3 shows the communication rates (words per minute) by LRM during a one-on-one conversation for 20 (N = 20) individuals using an AAC language application program supporting all three LRMs. The results show an average of a 282% (range 166% to 717%) communication rate advantage for multiple-meaning icons (semantic compaction) over spelling with no significant difference in communication rates between spelling and word prediction. These results demonstrate the importance of measuring these language components on an individual basis.
Finally, several language applications are based on similar LRM configurations, but do not have performance data available for consumers to make comparisons. These language applications may be installed on different user interfaces with different selection interfaces (as the secondary components described in the next section), further influencing performance and outcomes. For example, although the participants in the study represented in Fig. 3 were using different versions of the same language application program, the versions all supported access to the three LRMs with 128 locations. Current versions of this language application provide the choice of 144, 84, 60, or 45 locations. Other language application programs with other LRM availability offer different numbers of locations or visual scenes, but similar performance data have not been reported. Thus, software innovations have made gains beyond the pace that quantitative performance and outcomes data have been reported.
Advances to primary components
The software designs of the AAC systems have made significant advances in performance because of improvements to how language is represented and generated using technology. The primary components of AAC technology are related to how the system can perform the functions of a natural language ( Fig. 1 ). Having the features of language available is critical to achieving the goal of AAC and providing the most effective communication possible. Language performance is influenced by the availability of the three AAC language representation methods (LRMs), the selection and organization of vocabulary, and the method of constructing messages/utterances. These language-based components of the technology generally influence communication performance more than any other components.
In evaluating the full range and large variety of AAC systems available, all AAC systems support only up to three language representation methods (LRMs): (1) alphabet-based methods, (2) single-meaning picture symbols, and (3) multiple-meaning symbols. The three AAC LRMs may be available singularly or as a combination of methods simultaneously depending on the software program and appearance of the visual overlay or display. Each of the three LRMs can be defined, characterized, and identified as available or unavailable on a system. Briefly, each is defined in the following sections.
Alphabet-based Methods
These use traditional orthography (letter-by-letter) and acceleration techniques that require literacy skills. The user selects from a standard keyboard, alphabet array, or whole word overlay to enter text to generate a message. Word prediction, while not an acceleration technique, is included as an alphabet-based method.
Single-meaning Pictures
These employ use of graphic symbols to represent one word or message. The symbols may be photos, line drawings, color graphics, or animated. A large symbol set is required to represent the typical spoken vocabulary, but literacy skills are not required. One subset of single-meaning pictures is visual scenes.
Multiple-meaning Icons
The use of a small set of symbols to select vocabulary and linguistic structures to generate messages has been termed Minspeak. Icons on a single page are used in a prescribed sequence to access a large vocabulary. Literacy skills are not required.
Vocabulary is selected and organized on AAC systems based on vocabulary frequency and language development and use. Vocabulary studies have confirmed that the high-frequency words used by speakers are consistent across cohorts and are a relatively small pool of words. These core words consist of about 450 to 500 words, but make up approximately 80% to 85% of the words used in a language sample regardless of the topic. Extended vocabulary consists of thousands of words that only make up 15% to 20% of the words used in a conversation related to a specific topic or activity. Modern vocabulary databases provide AAC software engineers with high-frequency vocabulary lists to support the selection and organization of core vocabulary. The most efficient AAC systems provide quick and random access to high-frequency vocabulary to avoid spelling and the need to search for locations. Also, single-display vocabulary organization allows word access to become automatic as in touch typing.
Methods of utterance generation are the third language component that influences the performance with an AAC system. Word-by-word construction of messages, when used on AAC technology, provides for spontaneous, novel utterance generation (SNUG). Spontaneous generation of self-created utterances is the hallmark of human language and corresponds to how children develop language and adults use language in everyday activities for full participation. SNUG allows individuals to say exactly what they want and intend to say. The design considerations regarding availability of the AAC LRMs and organization of core vocabulary influence SNUG. The use of prestored messages provides an alternative to spontaneous self-generated text and communication, but has several important performance issues.
Advances in computer speed and memory have provided the means to incorporate sophisticated algorithms for predicting language units longer than single words. Preprogrammed phrase and utterance-based programs may aid literate individuals who use AAC to engage in social conversations and structured situations. As the term implies, the quick-hit or quick-fire features, depending on the product, support pragmatic language functions. These one-hit prestored messages allow individuals convenient and prompt selection of prestored social comments, interjections, polite remarks, and conversational fillers to provide for turn-taking and topic maintenance during conversations. Fig. 2 shows an AAC system that has, among its other communication modes, programming to support social language use by prestored utterances and scripts appropriate for specific topics and situations or general conversational turn taking.
In summary, today’s AAC language software is multifaceted with complex programming that provides for the characteristics of a natural language. The MPT process starts with prioritization of the primary language considerations by ensuring the individuals are informed fully of all available options and performance information. Rehabilitation clinicians, however, must be careful when appraising the evidence base. For example, several studies have shown that word prediction is not any faster than spelling. Yet, word prediction is referenced routinely as a rate enhancement strategy in some product literature and textbooks. Because word prediction reduces the number of key strokes for text entry, the feature still may be recommended for patients who fatigue with AAC technology use. Fig. 3 shows the communication rates (words per minute) by LRM during a one-on-one conversation for 20 (N = 20) individuals using an AAC language application program supporting all three LRMs. The results show an average of a 282% (range 166% to 717%) communication rate advantage for multiple-meaning icons (semantic compaction) over spelling with no significant difference in communication rates between spelling and word prediction. These results demonstrate the importance of measuring these language components on an individual basis.
Related posts:
Quality-of-Life Technology for People with Spinal Cord Injuries
Research in Computer Access Assessment and Intervention
Electronic Aids for Daily Living
Quality of Life Technology: The State of Personal Transportation
Virtual Coach Technology for Supporting Self-Care
Using Architecture and Technology to Promote Improved Quality of Life for Military Service Members with Traumatic Brain Injury
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
