RESNA 26th International Annual Confence

Technology & Disability: Research, Design, Practice & Policy

June 19 to June 23, 2003
Atlanta, Georgia


Sandra L. Hubbard, MA, OTR and Donald M. Spaeth, PhD
Department of Rehabilitation Science Technology, University of Pittsburgh


Computer technology can enhance the lives of individuals with disabilities if matched to the user's needs and abilities. Using a single-subject ABA design, this study investigated the use of a computer data-logging program to quantify rate, accuracy, and efficiency effects of individualized alternative computer access interventions for three subjects. A "text-timer" program was written to measure the characters per minute entered by three individuals with moderate to severe motor impairment. The standard keyboard/mouse configuration was the baseline and an alternative access configuration was the treatment. Data did provide evidence that alternative computer access increased the rates of all three participants, and the accuracy and efficiency of two participants.


Table 1 Results for mean rate, accuracy and efficiency




Participant 1

Participant 2

Participant 3








17.9 (2.98)


17.79 (2.97)

8.46 (1.41)

19.4 (3.23)


















0: 0

0: 0

1: 1


1: 3.07

1: 1.67

1: 1.65

Assistive technology including alternative computer access is being developed and sold at a rate faster than clinical benefits can be measured and documented. As a result, technology interventions in the rehabilitation arena are being implemented according to trial and error rather than according to evidence. Therapists in the field have recognized a need to move away from concentration on the technology per se, towards the development of effective interventions using the technology [1]. The assistive technology is not the end goal, but merely a means to an end [2]. Not only is accurate measurement of a user's abilities a key component of successful intervention in computer access [3], it is a key component in generating the quantitative performance data documenting the degree of benefit the user receives from the computer access intervention, thus the efficacy of the intervention itself.
Figure 1. Useful characters per minute over three phases for participant one.
Figure 1
A line graph of participant one's text entry rate with and without the intervention

Qualitative methods of measurement of AAC performance have been and continue to be developed. Hill and Romich developed the Language Activity Monitor (LAM) to gather data on AAC system performance in the natural environment, to enhance clinical decision making and objective outcomes measurement [4]. Lesher and colleagues are developing universal data logging and analysis tools in augmentative communication as a means of complex analyses of user-machine interactions [5]. Saunders and Saunders developed an automated data system for the measurement of frequency and duration of switch use, thus, increasing the ability to make informed decisions about the effectiveness of switch use [6, 7]. No automated measures of text entry into a computer were found.  

Figure 2. The ratio of raw to useful output over five trials for participant one.
Figure 2 A bar graph comparing participant one's total characters entered with the characters that were accurate

The purpose of this study was to determine whether a data-logging computer program can be used to measure the text entry rate, accuracy, and efficiency of individuals with moderate to severe motor impairments. This would allow the therapist and consumer to determine whether some alternative individualized computer access interventions are more efficient than the standard keyboard and mouse. 


A single subject, ABA design, was used with three participants having moderate to

Figure 3. Useful characters per minute over three phases for participant two.
A line graph of participant two's text entry rate with and without the intervention

severe motor impairments. During the baseline phase, participants entered text with a standard keyboard and mouse. During the second phase, participants entered text using one of three computer access modifications: voice recognition software, Minspeak® rate enhancement via Liberator™ or Pathfinder™ communication aid, and the Headmaster™ pointing


Using the "text-timer" program, optimum text entry rates of three individuals with moderate to severe motor impairments were measured over time. Their useable

Figure 4. The ratio of raw to useful output over four trials for participant two.
A bar graph comparing participant two's total characters entered with the characters that were accurate

text entry words per minute rate ranged from a mean of 1.41 to 3.23 words per minute (wpm), accuracy ranged from 44% (voice recognition) to 88% (Minspeak®), and efficiency ranged from 1hit in: 1 useful character out to 1hit in: 3.07 useful characters out. See Figures 1-6 and Table 1. 


Performance data, in addition to descriptive data, is necessary to document need and progress of therapeutic interventions (K. Hill,
Figure 5. Useful characters per minute over three phases for participant three.
A line graph of participant three's text entry rate with and without the intervention
personal communication, April 2002), [6]. The "Text Timer" program can quantitatively measure the performance (rate, accuracy, and efficiency) of individuals using alternative methods of computer access to enter text into a computer. The text entry rates of 1.41 to 3.23 wpm documented in this study are quite slow compared to 40 wpm of a non-disabled non-secretary typist, 22 wpm of an non-disabled individual who types random letters, and 9 wpm of a non-typist [8]. Even one word per minute is of clinical importance however, as it is a beginning that can be improved upon, and it is the difference between having and not having a social voice. Further, this study demonstrated that increased rate may be at the cost of efficiency: a method that yields a higher rate may also be less efficient. Thus both need to be taken into consideration when monitoring the effects of computer access interventions.

Figure 6. The ratio of raw to useful output over ten trials for participant three, using the standard and Minspeak® configurations.
A bar graph comparing participant three's total characters entered with the characters that were accurate


  1. R. Sevcik and M. Romski, "AAC: more than three decades of growth and development," The ASHA Leader, vol. 5, 2000.
  2. J. Treviranus, "Mastering alternative computer access: the role of understanding, trust and automaticity," Assitive Technology, vol. 6, pp. 26-41, 1994.
  3. H. Koester and W. McMillan, "Software for assessing computer usage skills," presented at Technology for the New Millennium, Orlando, 2000.
  4. K. Hill and B. Romich, "AAC best practice using Language Activity Monitoring," presented at CSUN, Los Angeles, 2000.
  5. G. Lesher, B. Moulton, G. Rinkus, and D. Higginbotham, "A universal logging format for AAC," presented at CSUN, Los Angeles, 2000.
  6. K. Hill and B. Romich, "A rate index for augmentative and alternative communication," International Journal of Speech Technology, vol. 5, pp. 57-64, 2002.
  7. M. Saunders and R. Saunders, "Automated data collection of microswitch use with persons with severe multiple impairments," presented at CSUN, Los Angeles, 2002.
  8. R. Lucky, Silicon dreams, 1st ed. NY: St. Martin's Press, 1989.

Sandra Hubbard, MA, OTR,
Human Engineering Research Laboratories (151R1)
7180 Highland Drive,
Pittsburgh, PA 15206
(412) 365-4850

RESNA Conference Logo