AAC Performance Based on Semantic Organization Schemes Using Dynamic Displays
ABSTRACT
This pilot study investigated the performance of two levels of a semantic organization scheme on augmentative and alternative communication (AAC) dynamic display systems. Language samples were collected using the Language Activity Monitor (LAM) from nine participants (ages 5 to 44 years) with complex communication needs (CCN). All participants had access to the three language representation methods (LRMs) of single meaning pictures, alphabet-based methods, and semantic compaction. One group of participants used a multi-level semantic organization scheme and a second group used a fixed-level scheme. Analysis revealed significant differences in the performance pattern to access vocabulary between the two conditions.
KEYWORDS
Language activity monitor, evidence-based practice, AAC, dynamic display
BACKGROUND
Augmentative and alternative communication (AAC) clinical decision-making is moving toward increased application of evidence-based practice (EBP). EBP expects practitioners to evaluate the external (field) evidence with the evidence collected at the personal level to make conscientious decisions about intervention (1). The introduction of language activity monitoring (LAM) has provided tools to support EBP and systematic methods to collect and analyze logfile data (1, 2).
Empirical data to support clinical decisions regarding the recommendation and implementation of AAC dynamic displays are limited, although a large number of such systems are available on the market. Clinicians are posing EBP questions about the usage patterns and communication competence of individuals using AAC dynamic displays. Clinicians are interested in identifying the advantages and disadvantages of dynamic display technology, also. Yet few studies are available that provide data on the semantic organization schemes to generate language using a dedicated AAC dynamic display system.
Three types of dynamic displays were examined in a single subject experiment that showed the most efficient performance for the fixed (all symbols appearing on one page) and the dynamic active (each symbol on one page linked to a second page) displays with the poorest performance was achieved with a dynamic passive (two pages of symbols linked by a “go to” symbol) display (3). However, children performed poorly in all three vocabulary scheme display conditions in another study (4). No data have been reported on the use of dynamic display systems with semantic organization schemes that include all three AAC language representation methods (LRM) of single meaning pictures, alphabet-based methods, and semantic compaction. These semantic organization schemes have been designed with the display configured for four core vocabulary rows that do not change and one activity row that changes with a total of 45 keys per display page. The core rows provide access to vocabulary stored under semantic compaction icons while the activity row and alternative pages provide access to vocabulary stored as single meaning picture symbols. The display schemes are either multi-level (dynamic active) or fixed-level (all symbols appearing on one page). The empirical assessment of comparing performance among various semantic organization schemes is critical to recommending AAC dynamic display technology in order to build communication competence in individuals with CCN.
RESEARCH QUESTION
The objective of this study was to investigate the performance of individuals who rely on AAC dynamic display systems. The specific research questions included: 1) What is the mean frequency of LRM use on multi-level and fixed-level display schemes? 2) Does a significant difference in the frequency of LRM use exist between individuals who use different display schemes?
METHOD
For this pilot study, nine individuals (N=9) with CCN were recruited. Participants were both male (N=5) and female (N=4), ranged in age from 5 years to 44 years, and included two individuals from a multicultural background. A primary diagnosis of cerebral palsy was used to identify five participants, while three participants had a primary diagnosis of mental retardation, and one participant had a syndrome not associated with cognitive challenges.
All participants had AAC dynamic display systems manufactured by the Prentke Romich Company that supported either a multi-level or fixed-level semantic organization scheme based on the Unity language application program. Six participants used a multi-level scheme on the Vanguard (N=4) or Vantage (N=2), and three participants used a fixed-level scheme on the Vanguard (N=1) or Vantage (N=2). Experience with the multi-level scheme ranged from five months to five years and on the fixed-level from one month to six months.
Participants were instructed to provide language samples representative of daily communication using their AAC systems. Language sample procedures were based on procedures designed as part of the SBIR research on the LAM (5). The built-in data logging feature of the Vanguard and Vantage was turned-on to record the language sample. Upon completion of the language sampling, the logfile was uploaded to a computer for analysis.
The primary researcher and one research assistant transcribed the language samples. The raw LAM logfiles were saved on a Gateway GP7-550 computer with Microsoft NT in a word processor. The logfile was then edited, coded, and analyzed using the Performance Report Tool (PeRT)(6). Reliability of the transcription of the data was determined through intra-judge and inter-judge agreement. The transcribers reached 100% agreement on sampling procedures before processing research data.
RESULTS
A total of 29 logfiles were collect from individuals who used a multi-level scheme for a total of 2662 utterances. A total of 46 logfiles were collected from individuals who use the fixed-level display for a total of 4569 utterances. Inter-judge reliability was assessed based on random sampling of 20% of the transcripts. The samples were compared for agreement on: 1) utterance segmentation, and 2) point-by-point agreement for codes. Random sampling reliability was 92%.
The transcript sets for all participants were analyzed to determine the frequency of use for the LRMs. For the purposes of this study core row activation was considered as use of semantic compaction and activity row or page use was considered use of single meaning pictures. None of the participants relied on alphabet-based methods. Table 1 shows the frequency of use for semantic compaction accessed using the core rows as well as the use of single-meaning pictures using the activity row or pages for both display schemes.
Organization Scheme |
Single Mean Pictures |
Semantic Compaction |
|---|---|---|
Multi |
56% |
43% |
Fixed |
35% |
66% |
Mean |
45% |
54% |
 |
|---|
A significant difference existed between the frequency of use for semantic compaction and single meaning pictures between the two organizational schemes. Individuals who used the multi-level display used single meaning pictures 56% of the time to access vocabulary while communicating. However, individuals who used the fixed-level scheme only used single meaning pictures with a frequency of 35%. Consequently, individuals who used the fixed-level scheme used the core rows significantly more frequently (66%) to access vocabulary. Individuals who used the multi-level scheme did not have as significant a difference between the use of semantic compaction and single meaning pictures.
DISCUSSION
This pilot study analyzed language samples to report preliminary performance patterns for two levels of a semantic organization scheme on an AAC dynamic display system. Using a display scheme based on core and activity rows should provide more immediate access to high frequency of occurrence (core) vocabulary stored using semantic compaction. Individuals who used the fixed-level scheme appear to be taking better advantage of the immediate access to core vocabulary provided by this display scheme. However, individuals with the multi-level scheme showed more use of multiple key activation to access picture producing vocabulary. Further research that looks at the frequency of core and extended vocabulary usage patterns that may exist among different organizational schemes is recommended. In addition, this research did not attempt to place controls on the inclusion criteria of the participants or control the communication intervention or environment. The primary goal of the study was to determine whether frequency usage patterns could be observed. With that goal achieved, future research is critical to contributing quantitative performance evidence to assist clinicians in making more informed decisions about the outcomes for AAC dynamic display technology.
REFERENCES
- Hill, K. J. & Romich, B. A. (2002). AAC evidence-based clinical practice: a model for success. AAC Institute Press. 2(1).
- Lesher, G., Moulton, B. J., Rinkus, G. & Higginbotham, D. J. (2000). A universal logging format for augmentative communication . In Proceedings of the 2000 CSUN conference , Los Angeles, CA: CSUN.
- Reichle, J., Dettling, E.E., Drager, K.D.R., & Leiter, A. (2000). Comparison of correct responses and response latency for fixed and dynamic displays: Performance of a learner with severe developmental disabilities. AAC: Augmentative and Alternative Communication, 16, 154-163.
- Drager, K.D.R., Light, J.C., Speltz, J.C., Fallon, K.A., & Jeffries, L.Z. (2003). The performance of typically developing 2-1/2-year-olds on dynamic display AAC technologies with different system layouts and language organizations. Journal of Speech, Language, and Hearing Research, 46, 298-312.
- Romich, B. A. & Hill, K. J. (2000). Language activity monitor feasibility study. National Institute for Deafness and Other Communication Disorders of the National Institutes of Health (NIH Grant No. 1 R43 DC 4246-01).
- Romich, B., Hill, K., Seagull, A., Ahmad, N., Strecker, J., Gotla, K. (2003). AAC Performance report tool. In Proceedings of the RESNA 2001 annual conference [CD-ROM]. Atlanta, GA: RESNA Press.
Author Contact Information:
Katya Hill, Ph.D., CCC-SLP,
Center for Assistive Technology Education and
Research (CATER),
102 Compton Hall,
Edinboro University of PA 16444,
Office
Phone (814) 732-2431
EMAIL: khill@edinboro.edu
