AAC Best Practice using Automated Language Activity Monitoring
Katya Hill
Edinboro University of Pennsylvania, Edinboro, PA USA
Barry Romich
University of Pittsburgh, Pittsburgh, PA USA
INTRODUCTION
Since the emergence of the field of AAC, professionals, family members, and consumers have agreed that the desire to communicate is the primary reason for using an AAC system. In fact, the goal for AAC is to provide the system and services that result in the highest level of personal achievement for the people who can benefit from the use of an AAC system (Hill & Romich, in press). Little research is available based on the actual communication of people who rely on AAC. Further, the AAC clinical intervention process seldom takes into consideration in an objective way the use of the AAC system between periodic therapy sessions. Conventional methods of monitoring AAC system use are based on personal observation and video or audio recording with subsequent observation, timing, and/or transcription. The cost of this approach is high because of the human time investment and the availability of the data is untimely. Consequently, professionals seldom collect data on the actual daily environmental use of AAC systems by consumers.Determining the effectiveness of clinical intervention requires evidence. With few, if any, exceptions, clinicians providing services to people who rely on AAC have not had the luxury of time necessary to obtain language samples from actual use of the AAC system in the natural environment. The obvious solution to this situation is the automation of the language data collection and analysis processes. There have been some efforts made in this area, all integral to specific communication or writing systems (Miller, et.al., 1990; Ahlsen & Stromqvist, 1999; Copestake and Flickinger; 1999). Some commercially available AAC devices have included limited features that monitor use, but until now none have incorporated time information.
LANGUAGE ACTIVITY MONITOR (LAM)
Romich and Hill (1999) are in the process of developing and testing the function of language activity monitoring. Work has been done on two configurations. Modern high performance AAC systems now include LAM as a standard built-in internal feature. For older AAC systems that have a serial port representation of what is spoken, an add-on LAM device is available from Prentke Romich Company. The LAM function records each language event (one or more letters, words, phrases, sentences, etc.) from the AAC system and attaches a time stamp. The data stored in the LAM over time is then periodically uploaded to a computer for editing and analysis. This uploading process can happen without custom computer software and the data can be transmitted as an email attachment or saved on a floppy disk for transmission or transport to a different location for analysis.COLLECTING LANGUAGE SAMPLES
Language samples can be gathered using various techniques in structured sessions. These include picture description, interview, narrative, and conversation. In addition, samples can be gathered in the long term (days or weeks) in the natural environment.PROCESSING RAW LAM DATA
The raw LAM data must be uploaded into a computer for editing and analysis. Presently, the HyperTerminal program which is part of Windows is being used for this. The use of a standardized reporting protocol provides for universal compatibility with computer-based applications (Hill & Romich, 1999). Below is an example of how the raw LAM data looks once uploaded into HyperTerminal:20:37:00 "I need "
20:37:05 "*[VOLUME UP] * "
20:37:06 "*[VOLUME UP]* "
20:37:07 "*[VOLUME UP]* "
20:37:14 "something "
20:37:16 "to drink "
20:37:19 "i"
20:37:20 "m"
20:37:24 "m"
20:37:28 "ediately "Once the raw LAM data is in HyperTerminal the text can be copied, pasted and edited in a word processing document. The final product is a language transcript that can be used for analysis. The above utterance example would look as follows after being edited:
I need something to drink immediately.
ANALYZING LAM DATA
After LAM data is uploaded into the computer and the raw data has been edited, the text is analyzed using standard and/or custom vocabulary and language analysis software applications. Various parameters to be analyzed could include:- vocabulary diversity
- frequency counts
- spelling versus whole word access
- morphological usage
- lexical or semantic usage
- syntactic structures
- developmental levels.LAM data analysis at the time of this writing has been based on the Systematic Analysis of Language Transcripts (SALT) (Miller, 1983). The time required to upload, edit, and analyze a two hour language sample of over 100 utterances from a high-end user takes approximately 15-20 minutes using these procedures. Simpler analysis, such as word searches, may take only a few minutes. The development of LAM tools to support editing and analysis will reduce analysis time significantly. Table 1 is based on SALT analysis of data from a two hour language sample collected from an intermediate user in the natural environment.
Table 1: SALT analysis
Total number of utterances 116 Mean length of utterance (MLU) 4.92 Number of different word roots 190 Total main body words 571 Type Token Ratio .33 To satisfy needs other than those addressed by standard language analysis programs it may be necessary to develop custom applications or use manual methods to analyze the LAM data. Examples of this could be analysis of use of particular vocabulary items since the previous therapy session, calculation of communication rate using the time stamps, methods used for language representation (single meaning pictures vs. spelling vs. semantic compaction), and error rates and types. Table 2 reports data collected from a high end user after two hours of recording in the natural environment.
Table 2: Language Representation Methods
Total word count 766 100% Total Minspeak words 707 92.3% Total whole messages (words) 1 (6) 0.8% Total single-meaning pictures 0 0% Total spelled & predicted words* 53 6.90% *Spelled words less anachronisms 51 6.64% Total spelled words 23 2.99% Total predicted words 30 3.90% Predicted / (spelled + predicted) 30/53 56.6% Total letter codes 0 0%
PRIVACY
One important issue in the use of any recording device is privacy. Both the LAM device and the internal LAM function provide for disabling the recording. It is strongly recommended that people whose communication is being monitored be clearly informed and that public use of recorded communication be anonymous. The LAM report header starts with the warning:*** CAUTION! ***
The following data represents personal communication.
Respect privacy accordingly.SUMMARY
Current best practice in AAC implementation emphasizes communication outcomes based on a team selecting outcomes from a functional curriculum model (Blackstone, 1990; Hill, 1996; Gray, 1998). AAC outcomes can be determined by noting positive changes in the attitudes of teachers, classmates, co-workers and others toward the consumer (Calculator, 1998). However, the tools described here provide more objective data to represent the quality and quantity of communication performance. Clinicians using LAM tools are able to document daily use of targeted vocabulary and language representation methods thus facilitating the intervention process across environments. AAC teams have the instrumentation to develop and monitor Individual Education Program (I.E.P.) goals and objectives and to qualify the implementation strategies and techniques used to facilitate AAC system use.CONTACTS
Khill@edinboro.edu
Bromich@aol.comACKNOWLEDGEMENT
The development of LAM tools has been supported by a grant to Prentke Romich Company from the National Institute for Deafness and Other Communication Disorders of NIH.REFERENCES
Ahlsen, E. & Stromqvist, S. (1999). ScriptLog: A tool for logging the writing process and its possible diagnostic use. Proceedings of the 1998 ISAAC Research Symposium (Dublin, Ireland, August 1998). John Clibbens, Filip Loncke, and Lyle Lloyd (editors). Whurr Publishers, London.Blackstone, S. (1990). In pursuit of opportunities and interdependence. Augmentative Communication News. 3:1-4.
Calculator, S.N. (1998). AAC and individuals with severe to profound disabilities. Handbook of Augmentative and Alternative Communication. S.L. Glennen, DeCoste, D.C. San Diego, Singular Publishing Group, Inc.
Copestake, A. & Flickinger, D. (1999). Evaluation of NLP technology for AAC using logged data. Proceedings of the 1998 ISAAC Research Symposium (Dublin, Ireland, August 1998). John Clibbens, Filip Loncke, and Lyle Lloyd (editors). Whurr Publishers, London.
DeRuyter, F. (1998). Concepts and rationale for accountability in assistive technology. Volume I: RESNA resource guide for assistive technology outcomes: measurement tools. RESNA.
Gray, S. (1998). AAC in the educational setting. Handbook of Augmentative and Alternative Communication. S.L. Glennen, DeCoste, D.C. San Diego, Singular Publishing Group, Inc.
Hill, K.J. (1996). Welcome to the real world using quality management tools to support the implementation process. Proceedings of the 10th U.S. Minspeak Conference.
Hill, K., & Romich, B., (1999) A proposed standard for AAC and writing system data logging for clinical intervention, outcomes measurement, and research. Proceedings of RESNA '99.
Hill, K., & Romich, B. (1999 in press) Augmentative communication: Empowering parents Part I: The goal, the team, and AAC rules of commitment. Exceptional Parent.
Kohn, J.G., LeBlanc, M. & Mortola, P. (1995). Measuring quality and performance of technology: results of a prospective monitoring program. Assistive Technology. 6:120-125.
Miller, J.F. & Chapman, R.S. Systematic Analysis of Language Transcripts (SALT). San Diego. College Hill Press.
Miller, L.J., Demasco, P.W., Elkins, R.A. (1990). Automatic data collection and analysis in an augmentative communication system. Proceedings of the 13th Annual RESNA Conference. Washington, DC 99-100.
Romich, B.A. & Hill, K.J. (1999) A language activity monitor for AAC and writing systems: Clinical intervention, outcomes measurement, and research. Proceedings of the RESNA '99 Annual Conference, Long Beach, CA. 19-21.