W3C home > Mailing lists > Public > w3c-wai-gl@w3.org > April to June 2000

FW: Telecom problems and design strategies for cognitive disabiliti=

From: Gregg Vanderheiden <gv@trace.wisc.edu>
Date: Fri, 12 May 2000 00:12:32 -0500
To: "GL - WAI Guidelines WG \(E-mail\)" <w3c-wai-gl@w3.org>
Message-ID: <D088364DDC78D211B9CA00104B978B860AAACA@nt.trace.wisc.edu>

>From the web page


      Annotated Bibliography and Research Recommendations

                      Chapter 1. Background
              Chapter 2. Guidelines and Strategies
         Chapter 3. Cognitive Abilities and Disabilities
            Chapter 4. Specific Cognitive Disabilities
 Chapter 5. Toward Better Guidelines and Strategies: Closing the
          Chapter 6. Recommendations for Future Research
                      Appendix 1: References
               Appendix 2: Annotated Bibliography

  This report is also available as a PDF download. In order to
 view and/or print the downloaded PDF file you will need to have
  the Adobe Acrobat Reader Software which you can get here. To
 open the compressed version you will need a program like WinZip
                          or StuffIt.
            * Telecommunications Report PDF [2.9MB]
 * Compressed version Telecommunications Report PDF.zip [1.7MB]


                        August 16, 1999

             Report and Research Recommendations by
                      Ellen Francik, Ph.D.

                    Annotated Bibliography by
                      Suzanne Levine, M.A.,
                     Shelley Tremain, Ph.D.,
                Ed Roberts Post Doctoral Fellow,
                        and Ellen Francik

                        Project Director
            Betsy Bayha, Director, Technology Policy
                  World Institute on Disability



 Research, publication and distribution of this document is made
 possible through the Rehabilitation Engineering Research Center
  on Universal Telecommunications Access, a partnership between
 Gallaudet University, the World Institute on Disability and the
    Trace Research and Development Center at the University of
    Wisconsin. Funding is provided by the U.S. Department of
  Education, National Institute on Disability and Rehabilitation
  Research Grant H133E50002. Opinions expressed in this document
        are not necessarily those of the funding agency.

  Additional funding for research was provided by the Ed Roberts
  Fellowship Program, a collaboration between the University of
California, Berkeley and the World Institute on Disability under
  a grant funded by the U.S. Department of Education, National
    Institute on Disability and Rehabilitation Research, Grant


      This report summarizes a literature review on use of
    telecommunications by people with cognitive disabilities.
   Section 1 of the report describes the research methodology.
Section 2 discusses general accessibility guidelines, identifies
 those that appear to be relevant to cognitive disabilities, and
    notes some of the barriers to usage they are designed to
      overcome. It outlines four classes of universal design
               strategies designers might employ:
      * Redundant, user-controlled modality of information.
 * Streamlined, user-controlled amount and rate of information.
                      * Procedural support.
                    * Content organization.

  Section 3 describes underlying cognitive abilities and notes
    some of the functions or user tasks which they may affect.
Section 4 reviews detailed findings from the literature for some
    specific disabilities: age-related cognitive disabilities,
  learning and language disabilities, brain injury, and mental
  retardation. Section 5 summarizes research that relies on very
  detailed analyses of tasks, cognitive abilities, and interface
                        design changes.

  Section 6 concludes by suggesting areas for further research.
    Better design leads to better guidelines, and so the first
  priority is the continued development of "reference designs,"
sample accessible telecommunications devices and services. Other
                   areas of research include:
            * Layered or streamlined functionality.
      * Training, procedural support, and problem-solving.
  * Better user descriptions to help designers envision actual
       product use by people with cognitive disabilities.
      * Increased attention to human performance modeling,
        developing parameters that represent people with
    * Correlative research, tying design to specific cognitive
   abilities, which will benefit the nondisabled population as
           well as those with cognitive disabilities.
      * A unifying framework, or "roadmap," to better link
     theoretical and practical work for designers' benefit.

The appendix of this report contains an annotated bibliography.


                          1. BACKGROUND

This report fulfills a deliverable of Rehabilitation Engineering
  Research Center on Universal Telecommunications Access by the
 World Institute on Disability (WID). It summarizes a literature
  review on use of telecommunications by people with cognitive

    This report discusses barriers to usage that people with
    cognitive disabilities experience with telecommunications
    products and services. It also discusses universal design
  strategies that designers might employ, and interface design

 It concludes with recommendations of studies that would lead to
        improved design and further guideline development.

     The methodology of the research is described below. The
    bibliography and annotations are in Appendices 1 and 2.

                       1 . 1 . METHODOLOGY

   This project began with a review of resource literature and
other print resources; to seek information about problems of and
      strategies used by people with cognitive disabilities.

    For this project, cognitive disability is defined as those
cognitive functional limitations associated with aging, learning
disabilities, mental retardation and traumatic brain injury. The
  functional limitations associated with these disabilities can
vary widely, from severe retardation, to memory difficulties, to
    the absence or impairment of specific cognitive functions,
  particularly language (Vanderheiden and Vanderheiden, 1991).
        They also may include difficulties in perception,
     problem-solving, conceptualizing, reading difficulties,
thinking, and sequencing (Trace Research and Development Center,
               1996d, EITAAC Final Report, 1999).

    Search terms used to describe these functional limitations
                          * Alzheimer's
         * Attention Deficit Disorder/Attention Deficit
                    * Hyperactivity Disorder
                     * Cognitive Disability
                           * Dementia
                   * Developmental Disability
                           * Dyslexia
                           * Epilepsy
                    * Intellectual Disability
                      * Learning Disability
                      * Mental Retardation
                        * Mental Handicap
                      * Stroke / Aneurysm
                    * Traumatic Brain Injury

    Telecommunications technology is broadly defined as those
  technologies requiring or supporting interaction with a person
    or information source. Examples would include equipment,
  software and interactive services such as telephones, pagers,
 Automatic Teller Machines, kiosks, and the Internet. Technology
            terms used in the search are as follows:
                          * Appliances
                  * Automated Voice Processing
                             * ATMs
                          * Caller ID
                        * Cellular Phones
                      * Computer Interfaces
                     * Consumer Electronics
                            * Pagers
                           * Internet
                      * Telecommunications
                          * Telephones
                          * Voice Mail
                        * Wireless Phones
                        * World Wide Web

  The study began with a list of key words related to cognitive
 disability, telecommunications and engineering design generated
    by project staff. The engineering design terms used were:
                            * Access
                     * Assistive Technology
                        * Cognitive Load
                       * Design Standards
                        * Human Factors
                        * Input / Output
                       * Interface Design
                * Interpretation of Information
                          * Navigation
                        * Physical Access
                       * Universal Design
                          * Way Finding

  From this list of terms, staff assembled a matrix of the terms
  to assist in database searches. An initial search using this
  matrix was conducted on-line through 14 different databases.
                      These databases were:
          * Applied Science and Technology Abstracts,
  * ArticleFirst, Online Computer Library Center, Inc., Dublin,
        * Business Dateline, UMI Provides, Ann Arbor, MI
  * Business and Industry, Responsive Database Services, Inc.,
                          Beachwood, OH
  * ContentsFirst, Online Computer Library Center, Inc., Dublin,
    * Educational Resources Information Center, Rockville, MD
         * Government Printing Office, Washington, D.C.
  * PapersFirst, Online Computer Library Center, Inc., Dublin,
            * INFO TRAC, health and diseases database
                 * MAG INDEX, magazine database
           * SocioAbs, Cambridge Scientific Abstracts
     * Social Science Abstracts, H.W. Wilson Co., Bronx, NY
     * Wilson Business Abstracts H.W. Wilson Co., Bronx, NY
    * WorldCat Database, Online Computer Library Center, Inc.,
                           Dublin, OH

 In addition to literature databases the World Wide Web (WWW, or
  Web) was used as a resource. Searches were conducted at more
 than 22 websites, including disability-related sites as well as
    some computer research sites. This proved to be fruitful -
 especially sites from conferences where presenters posted their
  papers on-line, as well as the Trace Research and Development
    Center site and Learning Disabilities on-line sites where
researchers have posted their work. The Web sites searched were:
      * Association for Computing Machinery - www.acm.org
       * California State University, Northridge Center on
                Disabilities -www.csun.edu/cod/
        * California State University, Northridge Papers -
      * Center for Accessible Technology - www.el.net/CAT
  * Center for Applied Special Technology (CAST) - www.cast.org
            * Closing the Gap - www.closingthegap.com
     * Disabilities Opportunities Internetworking Technology
            (DO-IT) - weber.u.washingtion.edu/~doit/
              * Epilepsy Foundation - www.efa.org
        * Job Accommodation Network - janweb.icdi.wvu.edu
  * LD OnLine: Learning Disabilities Information and Resources
      * Learning Disabilities Association - www.ldanatl.org
             * National Academy Press - www.nap.edu
    * National Information Center for Children and Youth with
                  Disabilities -www.nichcy.org
      * National Institute on Disability and Rehabilitation
          * Research - www.ed.gov/offices/osers/NIDRR
    * National Research and Development Centre for Welfare and
      Health (STAKES) - www.stakes.fi/cost219/COSB224.HTML
      * New England Medical Center. Department of Physical
  * Medicine and Rehabilitation Research and Training Center in
              Rehabilitation and Childhood Trauma.
  * Parents and Educators Resource Center - www.perc-schwabfdn.
   * Research Group on Rehabilitation and Independent Living -
    * TRACE Research and Development Center -trace. wisc.edu
       * Traumatic brain injury bibliography of web links.
                  * World Health Organization -
         * World Wide Web Consortium (W3C) - www.w3.org

The Job Accommodation Network (JAN) of the President's Committee
   on Employment of People with Disabilities was contacted for
      literature and case histories of people with cognitive
  disabilities requiring modification in the work place in order
              to use telecommunications technology.

 Shelley Tremain, the 1997 Ed Roberts Postdoctoral Fellow at WID
 and UC-Berkeley, visited the office of the Ontario Brain Injury
 Association (OBIA) in St. Catharines, Ontario, Canada, in order
to conduct a search of databases housed there, as well as OBIA's
                        reference library.

  Consultant Kevin O'Brien contributed several key articles from
  the literature on human-computer interaction, cognitive aging,
                  and individual differences.

                          1.2. RESULTS

  There are 111 documents in the bibliography. The documents can
                     be grouped as follows:
      * 23 discuss characteristics of people with cognitive
   disabilities or more general human cognitive capabilities.
   * 55 cover products currently used by people with cognitive
  disabilities or evaluate particular interfaces as a function
                      of cognitive ability.
          * 5 hypothesize uses of certain technologies.
    * 28 contain design guidelines (including cross-disability
          guidelines) or other organizing frameworks.

    More information was found on technology for people with
    learning disabilities than for people with aging-related
  cognitive disabilities. Technology for people with traumatic
    brain injury or mental retardation is even less frequently
 documented. Also, there is significantly more information about
   computers than about telephones or other telecommunications

                   1.2.1. User Characteristics

    The group of documents addressing user characteristics is
  discussed at length in Section 3. Of the documents not cited
there, several are worth mentioning. LaPlante and Carlson (1996)
summarize US survey data on prevalence and causes of disability.
    Estimates are provided for mental retardation and learning
    disabilities, but population estimates of other types of
    cognitive disabilities are hard to extricate from the data
  because categories are based on medical conditions that do not
  uniformly result in cognitive limitations. Brodin (1992) gives
 estimates of the prevalence of mental retardation in the world,
 Europe, and the US. She also defines some general requirements.
  The American Association on Mental Retardation (1999) defines
  mental retardation and provides other background at their Web
  site. Vanderheiden and Vanderheiden (1991) summarize cognitive
   functional disabilities as part of a larger survey. Several
       sources give definitions and statistics on learning
 disabilities, including Cramer and Ellis (1996), Hammill et al.
(1987), Johnson and Blalock (1987), Reid et al. (1991), and Shaw
                          et al. (1995).

    Articles on aging are discussed in section 4. Charness and
Bosman (1992) cover perceptual and physical as well as cognitive
aspects of aging. Park (1992) emphasizes applied cognitive aging
   research. West (1989) and Yesavage, Lapp, and Sheikh (1989)
      review the literature on mnemonic training and discuss
                underlying cognitive abilities.

           1.2.2. Product Use and Interface Evaluation

  A large group of documents covers assistive technology rather
  than universal design. Assistive technology refers to add-on
  devices or specialized equipment which provide access, whereas
    universal design describes access features that are fully
   integrated into off-the- shelf products. This group is also
  marked by the prevalence of documents about accommodations for
 learning disabilities. Surveys include those by Day and Edwards
  (1996), the Job Accommodation Network (1997c), Lewis (1998),
 MacArthur (1994), Raskind and Higgins (1998), Raskind, Higgins,
  and Herman (undated), Raskind and Scott (undated), Raskind and
                Shaw (1988), and Riviere (1996).

        More general surveys on assistive technology and
    telecommunications usage include those by the Alliance for
  Technology Access (1996), the Job Accommodation Network (1995;
  1997 a, b, d), and Vanderheiden (1998b). Mann (1996) discusses
    telephone accommodations for disabled elders and briefly
mentions memory dial buttons, including those with pictures, for
               people with cognitive disabilities.

  Bergman and colleagues (1991a, b; Bergman and Kemmerer, 1991;
  Bergman, 1998) provide short- and long-term evaluations of a
  computer environment customized for people with brain injury.
  Cole and Dehdashti (1998) summarize their work on a comparable
  system. Brodin and colleagues (1994; Brodin and Magnusson 1992
    a, b) examine the benefits of videophones for people with
                moderate retardation or aphasia.

  Glisky (1992; Glisky, Schacter, and Tulving 1986) demonstrates
  the benefits of an adaptive training system to teach computer
                  skills to people with aphasia.

    Individualized solutions - customized technology made to
  accommodate one particular person's disability - are presented
    by Chute (1988), Haugen (1996, 1997), Hendrix and Birkmire
 (1998), Hubble Dahlquist (1998), Kirsch et al. (1992), Kreutzer
    et al. (1989), Levinson (1997), and Steele et al. (1991).

  The remaining documents focus more on universal design - how
 easy mass market interfaces are for people of varying cognitive
 abilities. The Trace Research and Development Center has done a
series of grids analyzing problems that people with disabilities
  have with public phones, automatic teller machines, and public
    transport (1996 a, b, c). These analyses include dyslexia,
  language impairment, intellectual impairment and reduced motor
coordination. Trace has also published a reference design (1999)
  showing how a mass market wireless phone can be made compliant
              with Federal accessibility guidelines.

  Cress and colleagues (Cress and Tew, 1990; Cress, French, and
  Tew, 1991; Cress and French, 1992, 1994) analyze the cognitive
  complexity of computer input devices and evaluate their use by
  adults, normally developing children, and children with mental

Rogers and Fisk (1997) report difficulties older users have with
  automatic teller machines, and the effects of various forms of
    tutorials on users' performance and willingness to use the
  machines. Echt, Morrell, and Park (1998) also look at training
      materials when comparing young-old and old-old adults'
   acquisition of basic computer skills. Morrow and associates
    propose effective text-and-illustration formats for older
  adults' written medication instructions (Morrow et al., 1998b;
  Morrow and Leirer, in press). They also examine the effects of
  organization, enforced repetition, and optional repetition in
   automated appointment reminders delivered as phone messages
(Morrow et al., 1995; 1998b). Park (1992) includes other applied
        studies in her review of cognitive aging research.

  One set of these documents looks at the relationship between
   cognitive abilities and computer system design among people
 without disabilities. The seminal articles here are by Egan and
    Gomez (1985; Egan, 1988). They demonstrate the nuances of
 assessing individual differences, doing a sufficiently detailed
    task analysis to see where those differences matter, and
    redesigning the interface to minimize age, aptitude, and
  experience disadvantages. Sein et al. (1993) and Allen (1994)
 contribute other studies in this vein. Dillon and Watson (1996)
  provide historical background for this work. Anderson et al.
  (1995) take a different approach to accommodating individual
  differences by putting an adaptive user model at the heart of
                  their instructional software.

                1.2.3. Hypothetical Applications

       The third group of documents discusses hypothetical
     applications, that is, untested proposals of particular
 technologies to accommodate people with cognitive disabilities.
Three of these stand out. Barker (1998) describes general access
    strategies for WebTVT, and notes its flexible display and
ability to connect alternative input devices. Crandall, Bentzen,
  and Myers (1998) tested Talking Signsr with people with visual
 disabilities, and note the signs' potential use for people with
    cognitive disabilities who cannot read text. Vanderheiden,
 Mendenhall, and Andersen (1993) discuss the benefits of virtual
   reality for people who need help with written sequences, or
    people who need a more forgiving environment in which to
                      practice life skills.

                1.2.4. Guidelines and Frameworks

   The fourth group discusses guidelines and frameworks. Until
   recently, design guidelines for cognitive disabilities have
 tended to be general. Exceptions are Cress and Goltz (1989) and
  Vanderheiden and Vanderheiden (undated). The general usability
    standard from the American National Standards Institute,
      ANSI/HFES 200 (Blanchard, 1997), contains a section on
    accessibility but only in broad terms without reference to
 specific disabilities or specific modifications. The Electronic
Industries Association's guide (1996) is more specific about the
  design of consumer equipment. The World Wide Web Consortium's
Web Access Initiative (1998) provides guidelines for alternative
    visual and auditory displays, and proposes a style sheet
  standard. Nielsen (1996) discusses Web design for people with
    cognitive disabilities and Rose (1998) describes universal
 design strategies for K-12 curriculum development with specific
                 examples of cognitive supports.

 The efforts of the Telecommunications Access Advisory Committee
 (1997, 1999), and its successor, the Electronic and Information
  Technology Access Advisory Committee (1999) have provided more
      specific design guidance. The TAAC's working draft of
 accessibility guidelines and strategies (1999) is currently the
  most complete, product-independent list available. It compiles
              practical ideas for designers to try.

  The EITAAC report (1999) is notable for two things. First, it
  contains some technology-specific standards and implementation
 details. Second, it sets an accessibility goal. A person with a
disability should be able to "perform the same tasks, access the
same information, with the same approximate ease and in the same
   approximate time and at the same cost" as someone without a
     disability (4.3.1). This implies that designers should
  especially pursue solutions that help to close the gap in task
   performance between those with (cognitive) disabilities and
   those without. Vanderheiden and colleagues have developed a
 cross-disability accessible user interface framework for kiosks
     and other consumer devices (Law and Vanderheiden, 1998;
 Vanderheiden, 1997; Vanderheiden, 1998c; Vanderheiden, Law, and
    Kelso, 1998). The emerging Digital Talking Book standard
   (DeVendra, 1998; Kerscher and Hansson, 1998) which delivers
 books in a range of formats and a variety of navigation options
may also provide models for creating access to printed materials
             for people with cognitive disabilities.

      The use of a functional grid for product evaluation is
  demonstrated by Kaplan, De Witt, and Steyaert (1992) and the
      Trace Research and Development Center (1996 a, b, c).

  Shneiderman (1992) and Nielsen (1993) provide general design
    guidelines and principles for designing the computer user

 Two reports of the Human Capital Initiative provide a different
kind of framework: psychological research agendas. One is in the
 area of aging (Science Directorate of the APA, 1993) and one in
  productivity (Human Capital Initiative Coordinating Committee,
   1993). The report on aging emphasizes the need to study the
               oldest populations, those over 80.

                1.3. ORGANIZATION OF THIS REPORT

  How well do existing guidelines and design strategies capture
  the needs of people with cognitive disabilities? What research
                will lead to further improvements?

To answer these questions we need to be clear about four factors
      in the literature, and the relationships between them:
  * Distinct groups of people with cognitive disabilities: older
  adults, people with learning disabilities, people with brain
     injury, those with mental retardation, and so on. These
        distinctions should not be dismissed as diagnostic
      categories. To product managers, these terms represent
    markets, whose needs vary greatly. To design teams, they
     also represent distinct user groups to be recruited for
              assistance with product development.
   * Real-world tasks, or functions with which these different
      user groups may have difficulty, such as balancing a
      checkbook, keeping appointments, or remembering phone
  * Underlying cognitive factors that create those difficulties,
    such as problems with working memory, executive function,
            attention, or visual and spatial ability.
  * Specific design changes that compensate for those cognitive
    deficits. These evolve as technology evolves. New products
    may relieve some cognitive burdens and introduce others.

 There are many potential relationships between any two of these
factors. (For example, older adults and people with brain injury
  may both have difficulty keeping appointments, but differ from
 each other on other tasks.) It will be clear in this summary of
  the literature that different authors' frameworks clash, both
  within these factors and between them. Developing a unifying
  framework - call it a "roadmap" - is beyond the scope of this
  report but could be a research project. A roadmap would allow
  designers to trace a path from design change back through user
  task and cognitive factors to the populations served. Having a
  roadmap, designers could make more principled, organized, and
                  coordinated design decisions.

  Sections 2 through 5 of this report examine the relationships
  among these factors in increasing detail. Section 2 discusses
  guidelines for accessible design, identifies those that appear
 to be relevant to cognitive disabilities, and notes some of the
    functions or tasks to which they refer. Section 3 gives an
overview of cognitive factors and notes some of the functions or
  tasks to which they may contribute. Section 4 reviews detailed
  findings from the literature for some specific disabilities:
age-related cognitive disabilities, learning disabilities, brain
  injury, and mental retardation. Section 5 summarizes research
    that relies on very detailed analyses of tasks, cognitive
  factors, and interface design changes. Section 6 concludes by
    summarizing the state of cognitive disability design and
             suggesting areas for further research.


                    2. GUIDELINES & STRATEGIES

The most recent, and arguably most comprehensive, compilation of
  accessibility guidelines and design strategies is that of the
 Telecommunications Access Advisory Committee (TAAC, 1999). This
    is a working document maintained by the Trace Research and
  Development Center. It is separate from the committee's formal
  report to the Access Board in 1997. These guidelines, in turn,
  point to many guidelines and source documents that have come

   By the time design strategies show up in a cross-disability
  compilation like the TAAC guidelines, they have typically been
  abstracted away from particular disabilities. The document's
    organization is first in terms of engineering area (input,
   output, documentation, device compatibility), and second in
      terms of the user's functional task (understand spoken
  information, operate controls, insert disks or attach cables).

    Within the Input category of the TAAC guidelines, there is
    actually a subgroup of strategies specifically targeted to
cognitive disabilities. However, they are not further subdivided
  into strategies relevant to aging, mental retardation, brain
 injury, and so on. Other input strategies relevant to cognitive
 disabilities are found in "Availability of visual information,"
      "Access to moving text," and "Availability of auditory

  In the Output section there is no disability-specific advice,
   but again strategies relevant to cognitive disabilities are
  found under "Availability of visual information," "Access to
  moving text," and "Availability of auditory information," as
    well as "Prevention of visually-induced seizures." In the
 Information, Documentation, and Training section there are many
    recommendations on simplifying information presentation,
 complexity, and organization. The Compatibility and Specialized
      CPE section contains requirements affecting users of
       augmentative and alternative communication devices.

  Across the engineering categories of the TAAC guidelines, and
  throughout the rest of the literature - embodied especially in
    the case studies of accommodations - are four classes of
      * Redundant, user-controlled modality of information
  * Streamlined, user-controlled amount and rate of information
                      * Procedural support
                     * Content organization

  The first two classes of strategies benefit people with many
different disabilities, although the specific solutions may need
to be adjusted for cognitive disabilities. For example, the same
   accessible design that allows people with visual or hearing
   impairments to receive information in the modality of their
  choice (auditory or visual) also helps people with learning or
language disabilities. The third class of strategies, procedural
          support, is less well explored and documented.

 The fourth class of strategies, content organization, tends not
to be included in guideline documents that focus on products and
    services, except when documentation is discussed. Research
    literature on learning disabilities does discuss document
 content as well as reading tools. Increasing attention is being
         paid to accessible organization of Web content.

      The lists of detailed strategies are illustrative, not


These are compensation strategies frequently used by people with
learning disabilities or language disabilities (Day and Edwards,
  1996; Job Accommodation Network 1997c; Lewis, 1998; MacArthur,
  1994; Raskind and Higgins, 1998; Riviere, 1996). In addition,
  the ability to receive auditory information in static visual
     form is useful for people who need more time to process
 information (Brodin, 1992; Vanderheiden, 1997; Vanderheiden and
   Vanderheiden, 1991). Specific strategies mentioned in these
         sources and the TAAC guidelines (1999) include:
  * Use visual examples (diagrams, icons, drawings) in addition
                      to text descriptions.
  * Provide auditory (e.g., spoken) equivalents for all visual
  * Provide visual (e.g., written) equivalents for all auditory
      * Do both of these in a way that user may choose the
       presentation on demand; for example, request that a
    particular interface element be spoken aloud, instead of
                   requiring a screen reader.
              * However, do support screen readers.
        * Also support interfaces to alternative devices.
    * Describe pictures; caption or transcribe audio tracks;
              * provide audio description of video.
  * Use multiple methods to allow users to identify and locate
    controls: shape, size, texture, color, customized labels,
    customized keycaps, and/or voice output to announce keys.


    This class of strategies supports selective attention and
reduces perceptual and memory loads for people with all types of
  cognitive disability. It also compensates for hemineglect in
people with brain injury. Specific strategies come from the TAAC
    guidelines (1999) as well as work on learning disabilities
  (e.g., Day and Edwards, 1996; Raskind and Higgins, 1998) and
      brain injury (e.g., Bergman, 1991 a, b). They include:
  * Provide user control of size, placement, and appearance of
     display elements: high contrast, large print, numbered
  * bullets, strong highlighting, placement of elements on right
                  or left half of display, etc.
                  * Support Web style sheets.
  * Provide user control of pitch, volume, rate, and repetition
                    of auditory information.
            * Support screen magnification utilities.
        * Use simple screen layouts or one thing at a time
          * Use standard, simple layouts for controls.
   * Layer functionality; hide less frequently used functions;
      let the user customize the environment to foreground
    frequently used functions. (However, the risk is that some
    people may not look for "hidden" functions - out of sight,
                          out of mind.)
               * Make the product self adjusting.
     * Use orientation independent connectors and media. Use
    wireless connection strategies. Ensure that connectors and
              media cannot be inserted improperly.
  * Provide ways in which the user may recognize, rather than be
                required to recall, information.
  * Avoid flash or display refresh rates that induce seizure.
    * Keep all messages on the screen until the user dismisses
     * Provide a mechanism to speed up, slow down, or repeat
              information until it is acted upon.
    * Provide a non-time-dependent input method or make timing
      * Avoid functions that require simultaneous actions to
                      activate or operate.
    * Use a two-step "select and confirm" to reduce accidental
         selections, especially for critical functions.
  * Let the user set the pace of interaction with the system.
              * Reduce system lag and response time.

                     2.3. PROCEDURAL SUPPORT

   These strategies support executive functions, in particular
  planning and sequencing. They reduce memory load during long
   interactions with the product or system and help to counter
 distractibility. Integration with calendar or reminder programs
  can help to make a given telecommunication product or service
  part of a well-organized, productive day. Specific strategies
  are drawn from the TAAC guidelines (1999) as well as from work
  on mental retardation (Brodin, 1992), brain injury (Levinson,
  1995), and various cognitive disabilities (Job Accommodation
           Network, 1995; 1997 a, b, d). They include:
    * Structure tasks, cue sequences, and provide step-by-step
     * Provide definite feedback cues: visual, audio, and/or
     * Provide concrete rather than abstract indicators. Use
     absolute reference controls rather than relative ones.
          * Use goal/action structure for menu prompts.
  * Support "wizards" which offer help, simplify configuration,
                   and assist with sequences.
  * Automate complex sequences like system backup, application
                 launch, and user registration.
    * Provide defaults and make it easy to re-establish them.
    * Support integration with calendar or reminder programs.
     * Provide calculation assistance, or reduce the need to

                     2.4. CONTENT STRATEGIES

 This class of strategies is particularly useful for people with
language or learning disabilities (Day and Edwards, 1996; Lewis,
  1998; Raskind and Higgins, 1998). The literature on "content"
addresses primarily the creation and comprehension of text. This
section can certainly grow to include multimedia documents. Some
 of the issues related to multimedia documents have been covered
  in sections 2.1 and 2.2 above. Video captioning and Web style
         sheet support are examples of those strategies.

      By "content" we are also addressing not only product
     documentation (TAAC, 1997, 1999; EITAAC, 1999) but also
information services such as Web pages (Nielsen, 1996). Specific
            strategies cited in these sources include:
  * Provide a Web site map plus path information to the current
    * Structure text for easy scanning; provide headings. Use
        sequential numbers for numbered menus or lists.
      * Keep language as simple as possible. Highlight key
  * Use highly descriptive words as hypertext anchors. Avoid the
                     "click here" syndrome.
          * Search engines should support spell check.
    * Searches should support query by example and similarity
  * Users should be able to use word prediction and grammar and
       spell checkers in conjunction with all text entry.
      * Users should be able to use speech recognition as an
    alternative to text entry. Support speech recognition or
   speech note taking as well as writing or typing, or support
                interfaces to appropriate devices.

 The TAAC guidelines and strategies are fairly complete. Many of
   them are also measurable. A designer would know whether the
 product had been significantly improved: "automate procedures,"
 "use goal/action structure for menu prompts," "provide at least
        one mode that does not require a response time."

 But some of the strategies are a matter of degree. Without some
  form of user testing or human performance modeling, a designer
would not know to what extent people with cognitive disabilities
  actually benefit from overly generalized suggestions such as:
    "use simple screen layouts," "keep language as simple as
            possible," "reduce the number of choices."

      In fact, a designer might use several strategies, even
measurable ones, yet miss the combination that makes the product
  accessible to people with particular cognitive disabilities.
  Verifying accessibility depends on knowing the characteristics
  of distinct user groups, and assessing their performance with
  the complete product. These topics are discussed in more depth
                      in sections 4 and 5.



When designers want to know where guidelines and strategies come
  from, guideline writers often provide an overview of cognitive
  deficits or of tasks that people with cognitive disabilities
            find difficult. Here we will do the same.

      Frameworks that others have chosen may be drawn from
     neuroscience (Cole and Dehdashti, 1998), the individual
 differences literature in cognitive psychology (Carroll, 1993),
          or access engineering (Vanderheiden, 1998a).

Anyone interested in developing a roadmap that is strongly based
    in theory should consult Carroll's work. His comprehensive
     review and reanalysis of individual differences studies
demonstrate that eight broad cognitive abilities determine human
   performance. Narrower abilities are classified under one of
 these eight headings. To incorporate Carroll's framework into a
    roadmap that can be used by designers, significant work is
required to tie his organization more clearly to functional user
tasks and to interface design (see also Dillon and Watson, 1996).

 Here, we find it most expedient to list detailed findings under
  the high-level headings of Cole and Dehdashti. This should not
be taken as the ultimate taxonomy and placement of some items is
  quite tentative, since items gleaned from divergent sources in
  the literature may refer to user tasks, underlying cognitive
                abilities, or some blend of both.

    An important distinction that is not captured by Cole and
  Dehdashti's taxonomy is that between "fluid intelligence" and
  "crystallized intelligence." Fluid intelligence is the ability
   to develop techniques for solving new and unusual problems.
  Crystallized intelligence is the ability to bring previously
  acquired, often culturally defined, problem-solving methods to
bear on the current problem. They respond differently to several
  variables, most notably aging. Fluid intelligence - flexible
  innovation and the ability to learn novel things - decreases
    from early adulthood onward. Crystallized intelligence -
 experience - remains constant or even increases throughout most
                of the working years (Hunt, 1995).

 Another distinction not captured here is the difference between
"level" and "speed" factors (Carroll, 1993). The first has to do
  with the ability to do a task; the second, with the ability to
 do it quickly. These are worth distinguishing because they make
  different predictions about user performance. For example, a
    test of computer input devices showed that those that were
    slowest for successful users weren't necessarily the most
          difficult to master (Cress and French, 1994).

                     3.1. EXECUTIVE FUNCTION

  This includes diverse areas such as problem solving, planning,
 self-monitoring, task sequencing, prioritization, and cognitive
           * Select and execute plans (Levinson, 1995)
                  Describe goals or procedures
              Determine appropriate action sequences
                        Initiate activity
            Repair plans if interrupted or postponed
  Implement solution and evaluate outcome (Vanderheiden, 1998a)
          Trace Research and Development Center, 1996d)
    Maintain persistent effort until task done (and no longer)

                * Time management (Levinson, 1995)
                     Understand past, future
        Generate realistic schedules (Vanderheiden, 1998a)
            Keep track of time (Vanderheiden, 1998a)
      Perform scheduled activities within time constraints

                * Self-regulation (Levinson, 1995)
                Use feedback to control behavior
                 Inhibit inappropriate reactions
                          Follow rules
    Understand consequences of actions, especially long-term
                      (Vanderheiden, 1998a)

                           3.2. MEMORY

     This includes short-term, long-term, verbal and visual,
          procedural, declarative, and implicit memory.
        * Retain and recall, both short-term and long-term
      (Broderbund and Alliance for Technology Access, 1997)

        * Sequential - retain and recall series in order
  (Vanderheiden, 1998a; Broderbund and Alliance for Technology
                          Access, 1997)

  The following abilities have working memory and/or long-term
                        memory components.
            * Categorize, perceive category membership
      (Trace Research and Development Center, 1996d; Levine,
                  Horstmann, and Kirsch, 1992).

                      * See causal relations
         (Trace Research and Development Center, 1996d).

        * Understand analogical or metaphorical relations
              (Levine, Horstmann, and Kirsch, 1992).

                  * Understand abstract concepts
         (Trace Research and Development Center, 1996d).

      * Generalize and apply previously learned information
         (Trace Research and Development Center, 1996d).

                  3.3. ORIENTATION AND ATTENTION

  This includes freedom from distractibility, focused attention,
                      and divided attention.
  * Be aware of self, or self in environment; not getting lost
   (Levine, Horstmann, and Kirsch, 1992; Vanderheiden, 1998a).

                 * Be aware of passage of events
              (Levine, Horstmann, and Kirsch, 1992).

            * Spatial relations: know left from right
    (Levine, Horstmann, and Kirsch, 1992; Miller and Shreve,

        * Compensate for hemineglect due to brain injury
                    (Bergman, 1991 a and b).

    * Sustained attention - maintain effortful or deliberate
                    (Miller and Shreve, 1994).

   * Selective attention - attend anywhere while filtering out
        irrelevant stimuli. Filter out background noise,
            overwhelming input. Manage visual clutter
         (Miller and Shreve, 1994; Vanderheiden, 1998a).

    * Alternating attention - shift rapidly between competing
                  stimuli or lines of thought.

                  3.4 VISUAL-SPATIAL PROCESSING

  This includes perception and integration of visual information
 in space. The ability to construct mental pictures and use them
  in problem-solving is important for mathematics (Hunt, 1995).

  Some taxonomies distinguish calculation as its own category,
      since, pragmatically speaking, people with cognitive
    disabilities have problems with calculation and assistive
  technology can help (Vanderheiden, 1998a). Mathematical tasks
  can, however, be shown to include specific components such as
visualization, sequential reasoning, induction, and quantitative
                   reasoning (Carroll, 1993).
    * Discrimination - perceive differences and similarities
          (Trace Research and Development Center, 1996d)

            * Organize and interpret what's perceived
      (Broderbund and Alliance for Technology Access, 1997)

                  3.5 SENSORY-MOTOR PROCESSING

  Some taxonomies do not include motor processing as a cognitive
   skill, since speed of motor response may be due to physical
    factors. Other taxonomies put language skill with a motor
  component (such as writing or typing) underneath a "language"

      Non-linguistic auditory processing is included here.
    * Discrimination - perceive differences and similarities
          (Trace Research and Development Center, 1996d)

            * Organize and interpret what's perceived
      (Broderbund and Alliance for Technology Access, 1997)

                          3.6. LANGUAGE

  This includes expressive and receptive language, repetition,
                prosody, speech rate, and fluency.
                             * Speak
            Choose correct word and sentence structure
  (Vanderheiden, 1998a; Trace Research and Development Center,

                            * Listen
       Understand or hear certain sounds, syllables, words
  (Vanderheiden, 1998a; Trace Research and Development Center,

 * Write and type (Miller and Shreve, 1994; Vanderheiden, 1998a)
          Form letters, order them (vs. mirror writing)
                (World Health Organization, 1980).
                Type, spell, choose correct words
      (Broderbund and Alliance for Technology Access, 1997).
            Organize text, develop discourse structure
      (Broderbund and Alliance for Technology Access, 1997).

       * Read (Miller and Shreve, 1994; Trace Research and
                    Development Center, 1996d)
                  Read individual letters, words
      (Broderbund and Alliance for Technology Access, 1997)
                Comprehend text, draw conclusions
      (Broderbund and Alliance for Technology Access, 1997)
  Understand maps, signs, symbols, pictograms, simple diagrams
                          or schematics
                     (Vanderheiden, 1998a).

                          3.7. EMOTIONS

     This encompasses control of and expression of emotions,
    detection and understanding of emotions, and frustration
tolerance. This category is noted here, but is not considered as
            an issue in the remainder of this report.
         * Be sensitive to others' emotional expression
                   * Take others' perspective
      * Respond differently to situations based on context
    (Levine, Horstmann and Kirsch, 1992; Vanderheiden, 1998a)



  Designers need to know the population they are designing for,
 and who to recruit to help test or evaluate the result. Here we
  summarize information found in these articles about specific
  cognitive disabilities, the functional and cognitive factors
     associated with them, and the design implications. This
information varies in breadth and depth. The articles on age are
  the most specific about the match between design and specific
    cognitive abilities. These findings should generalize to
populations with learning disabilities, brain injury, and mental
    retardation in cases where the same underlying cognitive
                    abilities are affected.

                            4.1. AGE

      Age is not so much a cognitive impairment as it is a
    circumstance that correlates with cognitive impairments.
  Differences in task performance among older adults are still
  great, despite the average decline in cognitive function with
     age. Part of what pulls average performance down is the
   late-life onset of Alzheimer's disease and of dementia; the
    percentage of adults with each of these ailments increases
  sharply between ages 65 and 85 (Trace Research and Development
                         Center, 1996d).

    A useful segment of the literature focuses on finding the
    underlying cognitive mechanisms responsible for the "age"
  effect, correlating task performance with those abilities, and
   directing redesign to minimize differences in performance.

                    4.1.1. Human Capabilities

  Cognitive processes that decline with age include attention,
working memory, discourse comprehension, inference formation and
interpretation, and encoding and retrieval processes. Areas that
  decline less, or not at all, include semantic priming tasks,
  picture recognition, implicit memory, prospective memory, and
    highly practiced or expert behaviors. Explanations include
    age-related slowing in learning and performing tasks, plus
failure to inhibit - attention to irrelevant detail (Park, 1992).

  In the literature on mnemonic functioning, older adults most
  frequently report problems remembering names. They also have
 problems with immediate memory (including what one was about to
  do), conversations, and recalling recent events (West, 1992).

  Middle age (45-64), old age (65-74) and late old age (75+) are
distinctly different in patterns of cognitive abilities. Slowing
  in response time becomes "pronounced"; by 75+ learning rate is
  halved from the 20s and there are significant declines in the
  ability to do two tasks at once. Fluid intelligence declines
  before crystallized intelligence and both have declined by 75+
(Charness and Bosman, 1992). Adults in their 70s test 15% to 40%
 lower on various measures of intelligence or cognitive function
than adults in their 20s, even when timed test-taking is removed
    as a possible factor (Yesavage, Lapp, and Sheikh, 1992).

       Older adults have less efficient working memory and
    problem-solving skills for novel tasks. They also slow in
response time and may have an overconfidence in their ability to
  handle risky situations compared with middle-age adults. The
  latter is a metacognitive problem (Charness and Bosman, 1992).
      Instructions that require inferring, reorganizing, or
 integrating multiple sources of information place heavy demands
 on older users' working memory (Morrow and Leirer, in press). A
    similar pattern shows up in studies of work performance.
 Cognitive ability, particularly working memory, may be the best
       predictor of work performance when the worker is in
     "transition," acquiring new information or learning new
 procedures. "Maintenance" periods - which may extend over years
  of experience - make fewer cognitive demands (Murphy, 1989; in
  Park, 1992). Integration of information and complex decision
     components is more difficult for older management teams
  (Steufert et al., 1990; in Park, 1992). Older adults may not
  show deficits in performance if the new information, however
        complex, fits into already well-learned schemes.

    Perceptual speed, the speed at which mental operations are
    performed, is also considered fundamental to considering
  age-related differences and may even contribute to differences
in working memory (Salthouse, 1985; Salthouse and Babcock, 1991;
             both in Echt, Morrell, and Park, 1998).

     Several studies look at how well variance in underlying
   cognitive abilities explains "age" results. Age did predict
  recall of automated health appointment information, even when
 working memory and note-taking accuracy were taken into account
  (Morrow et al., 1998a). But age did not predict comprehension
of, or memory for, printed medication information when cognitive
   abilities (verbal, spatial, working memory) were taken into
    account (Morrow et al., 1998b). In a study comparing older
 adults' mastery of CD-ROM-based vs. printed training materials,
  spatial and visual working memory best predicted success with
      the CD-ROM materials. Perceptual speed best predicted
  performance with the printed manual (Echt, Morrell, and Park,

                    4.1.2. Design implications

 Pick user and task models with care.

 Studies of younger adults show that matching the user interface
  to the user's cognitive representation of the task increases
  performance; however, the mental models must be tuned to the
  given task and individual (Carroll and Reitman Olson, 1988).
  Charness and Bosman (1990) outlined parameters for predicting
   the performance of older adults, based on the well-received
  model of Card, Moran, and Newell (1983). The field needs less
  biased samples from which to derive these design parameters.
  Laboratory work also needs to be extended to real-world tasks
          (all citations in Charness and Bosman, 1992). Provide streamlined, user-controlled amount and
                      rate of information.

    Several strategies are important here. First, streamline
 functionality. Because older adults have less efficient working
 memory and problem-solving for novel tasks, they should benefit
    more from the "training wheels" approach of limiting the
 application's overall complexity (Carroll and Carrithers, 1984;
         both citations from Charness and Bosman, 1992).

  Provide extra time for training, for learning novel tasks, and
  for performing tasks. A summary of several studies of training
 techniques show no interactions with age for learning software.
 However, another body of work indicates that older workers need
  more time and more assistance to go through whatever training
     regimen is chosen. For word-processing studies the time
  difference in completing a self-paced tutorial was 1.5 to 2.5
    times slower. Response time also becomes much slower with
            advancing age (Charness and Bosman, 1992).

 Minimize distractions. Providing extra study time doesn't fully
 compensate for complex or badly organized material. Age-related
slowing is more severe for complex, ill-defined tasks, even when
  older adults have unlimited time to learn. Similarly, memory
training has the potential of improving basic, general cognitive
processes, but older adults tend to be sensitive to context - or
  to the wrong aspects of context - and may not generalize the
    training (Park, 1992; Yesavage, Lapp, and Sheikh, 1992).
 Use prompts for procedures and support

  In particular, expect and support reliance on external memory
  aids. In studies of automated appointment reminders, enforced
    message repetition reduced age differences in recall. When
 repetition was made optional, older patients did take advantage
 of it. (This indicates some metacognitive skills at work.) They
  paid enough attention to take accurate notes but not enough to
recall as much as younger patients (Morrow et al., 1998a; Morrow
  et al., 1995). External strategies - lists, calendars, visual
  cues, notes - are older adults' most common type of memory aid
  in everyday situations. They are preferred to internal memory
 strategies (West, 1992). A simple accommodation is the use of a
  phone with memory buttons for older adults who have difficulty
remembering numbers. Pictures can be added to the memory buttons
  for those who have trouble with names or codes (Mann, 1996).

     Support content strategies.

   Bear in mind that training format matters less than careful
design of materials. One study evaluating four types of training
  found that ATM accuracy and transaction times were best for an
    online tutorial group, intermediate for detailed text and
 pictorial guide groups, and worst for a description-only group.
   However, even with the tutorial, older adults were only 60%
  correct in completing all the components of each transaction
   (Rogers et al.; in Rogers and Fisk, 1997). However, a study
   comparing older adults' mastery of CD-ROM-based vs. printed
 computer training materials found no performance difference due
    to type of material. In this case, the printed manual was
  comparable to the CD-ROM in content except for animation, and
 the manual format was designed with older adults in mind: large
    sans serif print, concise content, bolded key terms, and
    easy-to-see colors (Echt, Morrell, and Park, 1998). Again,
   pictures are no panacea; those that require more inferences
          penalize older adults (Morrow et al., 1998b).

    Require fewer inferences, or help people to make them. In
  studies of medication information, iconic medication timelines
  were added to well-designed text instructions. When the icons
  were integrated (with more explicit information about time and
   dose, thereby requiring fewer inferences), both younger and
      older adults comprehended them better than text alone.
  Nonintegrated icons, which required more inferences, penalized
 older adults. Note that this means that including illustrations
in documentation is not automatically beneficial (Morrow et al.,
  1998b; compare the TAAC guidelines and strategies, 1999). In
  studies of wayfinding, it was easier for older adults to learn
     their way around unfamiliar environments when helped to
integrate different views or perspectives into a mental model of
                  the entire space (West, 1992).

Re-use familiar, well-learned organizational schemes. In studies
   of spatial memory, older adults using consistent, organized
 arrays for objects find them as easily as younger adults. Older
 adults benefit from familiar environments because of self-paced
  knowledge acquisition, distributed practice, and overlearning.
                          (West, 1992)

  Experiments on medication instructions showed that older and
   younger adults actually share an organizing scheme for such
     information. (This is culturally shared, "crystallized"
  intelligence, as described by Hunt, 1995. It is less likely to
decline with age.) Written instructions using this common scheme
  increased recall 13 - 20% overall. Additionally, a list format
  reduced age differences in comprehension time and particularly
  benefited the oldest (over 70 years) participants (Morrow and
                       Leirer, in press).


  Learning disability (LD) is a generic term covering multiple
   functional impairments. Because so many people with LD have
 problems with written and spoken language (Raskind and Higgins,
1998; Reid et al., 1991), accessibility guidelines often include
  people with language impairments in the same category as those
                            with LD.

                    4.2.1. Human Capabilities

Learning disabilities are marked by problems in the "acquisition
 and use of listening, speaking, reading, writing, reasoning, or
mathematical abilities" (Hammill et al., 1987). Individuals with
  LD will have significant difficulties in one or more of these
   areas but also relative strengths in others. By comparison,
 individuals with mental retardation will have a flatter, lower,
             "strength" profile (Shaw et al., 1995).

    The most commonly reported difficulties are in reading and
  writing, but it's likely that oral language and mathematical
   disabilities are underreported (Johnson and Blalock, 1987).
      Working memory seems to be implicated in problems with
    mathematical reasoning, since that disability becomes more
  apparent as cognitive load increases. Another study indicates
 that LD readers' working memory deficits are greater than those
    of nondisabled readers (Reid et al., 1991). Crystallized
    knowledge - accumulated experience - is also affected. LD
   students with reading problems frequently have slow lexical
  access, which decreases their attention and comprehension. As
 they grow older, this leads to cumulative deficits in knowledge
  and language acquisition (Snider and Tarver, 1987, in Reid et
                           al., 1991).

  Educators evaluating people with LD should be able to identify
 information processing difficulties that explain the functional
  impairments: short- or long-term memory, strategic learning,
automatization of skills and strategies, or executive control of
  strategies. Further limitations in sensory abilities, physical
 abilities, or psychosocial skills may occur with LD or may help
rule it out in favor of another explanation (Shaw et al., 1995).

                    4.2.2. Design Implications Provide redundant, user-controlled modality of

  The same accessible design that allows people with visual or
  hearing impairments to receive information in the modality of
   their choice (auditory or visual) also benefits people with
  learning or language disabilities (Day and Edwards, 1996; Job
    Accommodation Network 1997c; Lewis, 1998; MacArthur, 1994;
    Raskind and Higgins, 1998; Riviere, 1996). People who have
  problems with written text can hear it spoken, and vice versa.
In one typical study, people with dyslexia who were able to scan
  written materials into an optical character recognition device
      with speech synthesis increased their reading rate and
  comprehension (Elkind, Black, and Murray, 1996, in Raskind and
     Higgins, 1998). In another, adding speech to text had a
  rehabilitative effect: it doubled the rate at which students
  acquired decoding (reading) skills (Wise and Olson, 1994, in
                          Lewis, 1998). Provide streamlined, user-controlled amount and
                      rate of information.

  Features particularly useful to people with LD are the ability
   to speed up, slow down, or repeat information, particularly
                      auditory information.

  Related to this is the ability to focus attention and minimize
  distraction by having the product display information in large
  print or high contrast, and by highlighting text by the word,
    phrase, or sentence as it is read aloud (Job Accommodation
  Network, 1997c). Distraction is also a potential problem with
  multimedia enhancements to static text. Hypermedia in talking
 books is intended to provide pronunciation aids, illustrations,
   and other reading aids. Its secondary benefit is to promote
  engagement with the text (Lewis, 1998; MacArthur, 1994). But
  animation and other "hot" visual effects can be distracting.
  Swanson (1989, cited in Lewis, 1998) argues that individuals
  with learning disabilities are typically "inefficient rather
    than passive" learners. If so, animation will draw their
  attention away from, rather than engaging their attention to,
                            the task. Prompt procedures and support decision-making - but
       don't expect everyone to use very structured tools.

People with learning disabilities use organizing devices ranging
 from "freeform databases" (simple electronic lists), to watches
  with alarms, to voice activated day planners with voice input
  technology (Job Accommodation Network, 1997c; Riviere, 1996).
      From this range of devices one can begin to infer the
  variability in users' needs. Some need text input and output,
   and some need spoken language. Some benefit from structured
 tools, while others need to capture information quickly without
      having to worry about how to enter it into the system.

 Task specific planning and decision aids are also useful. Since
    so much of LD accommodation focuses on language, the most
  commonly cited planning and decision aids are outliners, idea
    generation and capture programs, and other writing tools.
    Students with LD typically have problems with the planning
  activities associated with writing: setting goals, generating
  content through memory search and information gathering, and
    organizing material carefully (Flower and Hayes, 1981, in
                        MacArthur, 1994).

    Talking calculators and other mathematical tools are also
         included in the "procedural support" category.

     Support content strategies.

  All the strategies listed in section 2.4 apply when designing
 for people with learning disabilities. These include the use of
well-structured text, aids to text entry, and support for speech
  recognition and speech notetaking. The speech technologies are
          especially appropriate for portable, networked
  telecommunications devices. Voice dialing is an alternative to
number entry. Telephone recording (as with an answering machine)
   lets people review the content of a phone call at their own
 pace. Speech notetaking is an alternative to written notes (Job
                 Accommodation Network, 1997d).

                   4.3. TRAUMATIC BRAIN INJURY

  People whose cognitive functioning has been affected by closed
   or open head injury or by stroke are said to have traumatic
brain injury (TBI). People with TBI vary greatly in severity and
    type of cognitive disability. The literature on "cognitive
   prostheses" - computer systems that support these people in
  resuming activities of daily living such as managing household
    affairs - emphasizes highly customized software with few
    functions (Bergman, 1991a, b; Bergman and Kemmerer, 1991;
  Bergman, 1998; Chute et al., 1988; Cole and Dehdashti, 1998).
  However, it appears that the people willing to spend time and
  money on a cognitive orthotic are those whose disability is so
      severe that they cannot use off-the shelf products.

                    4.3.1. Human Capabilities

  Brain injuries often result in selective, rather than global,
  memory deficits. For example, amnesic patients can acquire a
  variety of motor, cognitive, and perceptual skills in a nearly
  normal fashion but not remember the learning episodes (Glisky,
1992). Cognitive dimensions of profound disability often contain
small pockets of ability, which are made visible during detailed
 software development of assistive technology (Cole & Dehdashti,

  Case studies illustrate how cognitive disabilities may cluster
and vary from individual to individual. Here are a few examples:
  * Difficulty and strain in attending to items on one side of
    the visual field (hemineglect); problems with short-term
  memory, sequencing, and executive functions (Bergman, 1991a).

  * Problems with memory, oral communication, processing speed,
   and concentration (Bergman, 1998). n Amnesia; also problems
        with prospective memory, attention, concentration,
  visual-spatial processing, and executive functions (Cole and
                        Dehdashti, 1998).

       * Expressive aphasia (communication limited to head
  movements); also memory limits and visual, spatial, or motor
   problems with computer cursor control (Chute et al., 1998).

  * Moving information from short-term to long-term memory: an
    inability to remember any new information after 30 minutes
                   (Cole and Dehdashti, 1998).

   * Short-term memory loss, auditory discrimination problems,
    and visual difficulties with computer screen flicker and
            glare (Job Accommodation Network, 1997a).

With appropriate accommodations and assistive technology, people
    who had spent years attempting to use computers and other
     communication aids have been able to live and work more
  independently. They were able to manage their own schedules,
  create and read text documents, balance their checkbooks, and
    pay bills. In one mild case, the remedy was as simple as
  providing a distinctive ringer on the phone and an anti-glare
    screen on the computer (Job Accommodation Network, 1997a).

     In the literature surveyed, only the work of Glisky and
  colleagues (Glisky, 1992; Glisky, Schachter, & Tulving, 1986)
    measures brain injured participants' underlying cognitive
abilities and correlates them with task performance. People with
   amnesia have lower scores on memory tests - for example, on
    delayed tests of logical memory and visual reproduction.
     However, people with mild or moderate amnesia, given an
  appropriate training environment, may equal nondisabled people
 in learning and performance of simple, structured tasks such as
                           data entry.

    On the average, amnesic patients learning basic computer
   programming and file manipulation started out with the same
   knowledge as non-disabled participants, eventually made few
    errors, and retained information over one or more months.
However, they took much longer to reach criterion. They did less
  within-session learning and more between-session forgetting.
    They also had difficulty answering general or open-ended
   questions about what they had learned (Glisky, Schachter, &
 Tulving, 1986). Amnesic patients learning data entry, a simpler
  task, could demonstrate knowledge they had gained even if they
 could not describe it. Their rate of gain in task speed was the
 same as that of non-disabled participants: 50% during the first
      training round, 15% during the second (Glisky, 1992).

                    4.3.2. Design Implications Provide streamlined, user-controlled amount and
                      rate of information.

  Most importantly, provide a mode with minimum functionality.
  Eliminate or hide what isn't essential. A Business Week cover
   story argues the benefits of streamlining consumer products
 (Nussbaum and Neff, 1991). The cognitive aids described here go
  much further (Bergman, 1991a, b; Bergman and Kemmerer, 1991;
            Bergman, 1998; Cole and Dehdashti, 1998).

  The design approach reads like a stringent version of the TAAC
  strategies (1999). Applications each had very few functions,
    compared to off-the-shelf software. (For example, the text
processor described by Bergman, 1991b, had two functions and was
   upgraded to three.) Error messages were concrete. Important
functions were moved to labeled, unshifted keys. Hemineglect was
  accommodated, for example, by shifting screen objects to one
  side of the display. Sequences were simplified, prompted, or
  eliminated. Redundant, multisensory cues were added. Typical
     components in both systems included a text processor, a
 check-writing program, a daily schedule, a paint program, and a
 remote control application that enabled the user to work with a
   therapist on line. In one case a scheduler with a pager was
     customized for a physician who sent himself appointment
information and other timed reminders (Cole and Dehdashti, 1998).

   The telecommunications equivalent of such a system might be
    something like the hypothetical "Companion" described by
  Vanderheiden (1992), a simple wireless phone with voice output
    and speech recognition, problem-solving aids, and reminder

  Also, support radical customization, either by the user or by
    someone working with the user. Both Bergman and Cole and
      Dehdashti report adding, deleting, or modifying entire
components of the product for each user. Cole and Dehdashti note
   that customization was instance-specific as well: cognitive
chunks in different parts of the software needed to be different
      sizes. Iterative, participatory design sessions were
  particularly common with early users of their system. In later
    versions they succeeded in doing mass customization more
                            quickly. Provide self-paced training and consider an
                        adaptive trainer.

Adaptive trainers are discussed further in section 5.3. Briefly,
  they block certain errors, diagnose others, suggest effective
    responses, and fade into the background as users' skills
    increase. Effective computer training systems for amnesic
   patients (Glisky, Schachter, & Tulving, 1986; Glisky, 1992)
   acted like adaptive trainers. They provided vanishing cues,
    evaluated responses, and occasionally offered more direct
  prompts. The systems were self-paced. It may also be important
      that users had the opportunity to learn the knowledge
              completely, or even "overlearn" it.
 Use prompts for procedures and support

  People with brain injury and those with learning disabilities
   have many of the same needs in this area. People with brain
    injury may be more likely to need support with executive
functions such as planning and scheduling. An interesting device
  that meets this need is PEAT, a hand-held electronic calendar
 and address book that has built-in telephone, fax, and Internet
 capabilities (Levinson, 1997). It generates daily plans, checks
     with the user about how activities are progressing, and
  reschedules activities in order of importance if the user is
falling behind schedule. The planner is based on a simulation of
              the brain's own planning functions.

                     4.4. MENTAL RETARDATION

     Mental retardation (MR) appears before adulthood. It is
     characterized by "significantly subaverage intellectual
   functioning" with related limitations in two or more of the
  following areas: communication, self-care, home living, social
    skills, community use, self-direction, health and safety,
functional academics, leisure, or work (American Association for
  Mental Retardation, 1999). Mental retardation is characterized
    by limitation in both intellectual and adaptive skills.

                    4.4.1. Human Capabilities

  People who are mildly retarded (80-85% of the population with
MR) have an IQ between 55 and 69. They achieve a 4th through 7th
grade education and function well in the community, holding down
 semi-skilled and unskilled jobs (Trace Research and Development
                         Center, 1996d).

    People with MR have working memory deficits. Perceiving,
 processing, and storing information require greater effort, and
  take more time. Specifically, people with mental retardation
  have problems with sensory discrimination. Also, since people
  with MR have difficulty thinking in abstract terms, concrete
  operations and a concrete environment are essential for their
                  accommodation (Brodin, 1992)

Note how the profile of people with MR is different from that of
people with learning disabilities or brain injury. In the latter
    cases there tend to be specific deficits but also relative
  strengths in many other functional areas. We would expect that
   some of the tools used by people with learning disabilities
      would need to be adapted for use by people with mental
  retardation. For example, a program that speaks written text
   would need to be much simpler than a typical computer-based
                          screen reader.

 The performance of people with mental retardation is not just a
  slower, scaled-down version of that of nondisabled adults, nor
    is it equivalent to that of nondisabled children. This is
  particularly true for tasks that have both cognitive and motor
  components. A person with mental retardation may be physically
    more capable of doing the task than a young person of the
      equivalent mental age, but may have greater cognitive

    In a study comparing learning of 5 input devices, adults,
children with MR, and normally developing children had different
 patterns of device mastery and task speed. The touchscreen (the
  fastest device overall) was significantly slower for children
      with MR, possibly because they tended to separate the
  simultaneous pickup and drag of targets into separate actions.
     Normally developing children mastered all devices after
  relatively brief training sessions; children with MR did not
    master the devices as easily, even with training, and had
  particular problems with the trackball and locking trackball
                    (Cress & French, 1994).

                    4.4.2. Design Implications Provide streamlined, user-controlled amount and
                      rate of information.

To do this, provide a mode with minimum functionality. Eliminate
  or hide what isn't essential. The "Companion" hypothesized by
  Vanderheiden and mentioned in the section on brain injury was
  actually created with mental retardation in mind. In terms of
 other strategies, user-defined pacing and adaptive training are
likely to be useful, but there is no direct evidence for that in
  these articles. Minimizing distractions is another promising
  strategy, but again there is no evidence here specific to MR. Provide redundant, user-controlled modality of

      Providing that operation is simple, people with mental
  retardation benefit from having text information spoken aloud.
They would also benefit from having status tones translated into
  something more meaningful, such as visual indication of call
 status (Kaplan, DeWitt, and Steyaert, 1992). In several studies
 of moderately retarded adults and children, the introduction of
    still picture phones increased callers' independent use of
      telephones. Communication was more effective and user
  satisfaction and self-confidence were greater, perhaps because
  these devices make visual and auditory information available
 simultaneously (Brodin and Magnusson, 1992 a, b; Brodin, 1994).
  Videophones with Pictogram monitors and document cameras were
  more complex and presented usability problems (Brodin, 1994). Prompt procedures and support decision-making.

    Vanderheiden (1992, 1998b) recommends such aids. The Job
    Accommodation Network (1997a) gives a low-tech example: a
  greenhouse worker carrying a tape recorder with reminders to
    stay on task and indicators of break time. Again, talking
 calculators, automatic checkbooks, and other mathematical tools
are included in the "procedural support" category (Vanderheiden,
   1992). Orientation aids are important for those who tend to
    wander or who may need to be located by emergency services
  (Vanderheiden, 1992, 1998b); this could equally well apply to
           older adults with dementia or Alzheimer's.

     Support content strategies.

  There is some design guidance on using simple language, if we
    take as a starting point that people with moderate mental
     retardation achieve a 4th through 7th grade education.



  Great improvements in product design can be made by providing
    access where none existed before. Provide information in
 redundant modalities; remove time-dependent controls; eliminate
sequences. Each of these strategies makes it possible for people
  with disabilities to use products they could never use before.

 But if designers take as their goal the EITAAC (1999) benchmark
  that people with disabilities should be able to "perform the
      same tasks, access the same information, with the same
  approximate ease and in the same approximate time and at the
    same cost," then a finer analysis of user ability and task
                      performance is needed.

  Individual differences in task performance, even among people
without disabilities, are massive. The ratio between slowest and
 fastest computer task times can be 20:1. Even the ratio between
  people at the 25th and 75th percentiles can be 2:1 to 4:1. For
  word processing and information search, the greatest source of
  differences is time in making and recovering from errors. In
programming, mental representation of the problem and the system
           also has a significant effect (Egan, 1988).

  Egan demonstrates that careful attention to these individual
    differences can guide interface design, resulting in the
 elimination of age, aptitude, and experience disadvantages. His
  goal is to aid the people having the most difficulty, without
  penalizing those with better performance. His methods work for
the cognitive differences underlying "age," and they should work
     equally well for differences underlying other cognitive


  The methodology for closing the gap is elucidated by Egan and
Gomez in studies of text editing performance (1985; Egan, 1988).
    First, assay individual differences. Find out which ones
  correlate with performance on the chosen task. Many plausible
  factors may be irrelevant. Egan and Gomez found that only age
  and spatial memory mattered for the text editors they studied;
    education, typing speed, verbal aptitude, job experience,
    associative memory, and logical reasoning had no effect on

     Second, isolate task components. Do a task analysis and
  determine which components individual differences affect. This
 is admittedly an art. Egan and Gomez verified their analysis by
  creating new tasks with the same components and obtaining the
    same pattern of correlations, but this additional test is
                  impractical for design teams.

 Third, accommodate the differences by redesigning the interface
  to minimize performance differences. Measure user performance
 again. Egan and Gomez compared a line editor to a newer display
editor. The newer design simplified command syntax, reducing age
  differences in performance, but the changes related to spatial
    memory ability were mixed (and cancelled each other out).

                      5.2. USER PROTOTYPES

  Egan (1988) outlined four ways in which user interfaces could
 accommodate individual differences. These approaches range from
          user interface focused to training focused:
    * Interface robustness: the improved mass market design of
    products used casually by large numbers of people. This is
          the approach discussed in section 5.1 above.

  * User prototypes: tailoring the interface to meet the needs
                of a few broad classes of people.

  * Adaptive trainer systems: supporting novice performance by
  prohibiting certain errors, diagnosing others, and prompting
    effective responses. Appropriate to settings where people
      need to learn a moderate amount to become productive.

  * Automated mastery learning: providing customized intensive
    training. Most appropriate where people need to master a
      great deal of material to a high level of proficiency.

  In practice, universal design uses both "interface robustness"
 and "user prototypes." Examples of the latter include providing
interfaces in different languages, or creating a basic mode that
  presents much less functionality than the full product. The EZ
  Access strategies (Law and Vanderheiden, 1998; Vanderheiden,
  1997; Vanderheiden, Law, and Kelso, 1998) could also be placed
                in the user-prototype category.


 Adaptive trainer systems would seem to be ideal for people with
   cognitive disabilities. Glisky (1992; Glisky, Schacter, and
  Tulving, 1986) used such a self-paced system to train people
  with amnesia on computers, using cues that gradually vanished
when no longer needed. Older adults may also benefit from fading
 cues that help them generalize to full, unaided use of a system
    (West, 1989). The concept is also known as "scaffolding."
Scaffolding is part of a universal design approach to curriculum
  which includes summarizing main ideas, providing scaffolds for
 learning and generalization, building fluency through practice,
        and assessing background knowledge (Rose, 1988).

  Adaptive trainer systems relieve memory burdens by providing
 assistance with novel sequences and prompting data entry (TAAC,

  As when designing more robust, universal interfaces, careful
  task analysis is essential. To design a trainer system, it is
  crucial to know how errors cluster, what their causes are, and
  what will promote transfer from the trainer system to the full
                      system (Egan, 1988).

 The line between trainer system and interface enhancement blurs
  in the case of "wizards," which remain available to users even
  after initial training. Wizards are included in today's mass
market products more frequently than Egan might have anticipated
                            in 1988.

    Automated mastery learning environments, although they are
  effective cognitive aids, are expensive to develop and require
   detailed user modeling. The self-paced math and programming
tutors developed by Anderson and colleagues (1995) are excellent
    examples. The cognitive benefits of using the tutors were
significant: in the best case, nondisabled students cut learning
  time by one- to two-thirds and scored significantly higher on
  post-tests. The success of that tutor was largely due to the
        sophistication of the user model, which predicted
    moment-by-moment what knowledge each user had or had not
  acquired. However, Anderson et al.'s other principles of tutor
 design can be taken from that report and added to accessibility
  guidelines and strategies. Some of them are already familiar:
    * Provide instruction in the problem-solving context - but
          between problems, not during them. Otherwise,
    problem-solving is interfered with. (This appears to be a
      case of interference with working memory and perhaps
                      executive function.)

    * Promote an abstract understanding of the problem-solving
  knowledge. Help students generalize principles from concrete

       * Minimize working memory load. Minimize irrelevant
    information and teach new components only when others have
                 been relatively well mastered.

    * Provide immediate feedback on errors. This is best for
    reducing task time, most likely by reducing time spent in
      error (see also Egan, 1988). Simply flagging errors or
     providing feedback only on request is less successful.

   * Facilitate successive approximations to the target skill.
      Let the tutor fade gradually into the background with
                       repeated practice.



     Functional guidelines already exist, as do a variety of
 practical strategies for making better products for people with
   cognitive disabilities. In the belief that product-specific
strategies will continue to be improved by design experience, we
  begin this section with the most concrete areas for research.

                      6.1. REFERENCE DESIGNS

   Designers often work from examples. A research project that
creates one or more examples serves several important functions.
 First, it demonstrates the feasibility of universally designing
appealing products that are also usable by people with cognitive
   disabilities. This is especially important given the stigma
associated with cognitive disabilities. Second, the examples can
  be publicized and specific design solutions can begin to find
 their way into commercial products. Third, working, user-tested
  designs create more data from which detailed strategies can be

 The Trace Research and Development Center has created just such
  a "reference design" for an accessible wireless phone (1999).
   Features in the reference design that are most relevant for
                   cognitive disabilities are:
  * An EZ button that allows one to have the label for any key
    read aloud, as well as the contents of the display and all
                menus and features of the phone.

     * An infrared port for wireless connection to assistive

      * A "one button feature" that allows the phone to be
      programmed so that it will automatically dial only one
    number (or a small list of numbers) when any keypad button
    is pushed. An optional cover plate to create a true "one
                   button phone" can be added.

     This is a good start for basic telephony functions. The
reference design doesn't address the complexity of the menus - a
  service-dependent issue, rather than a hardware issue, that is
  nonetheless an important one for users. Nor does it explicitly
  address the degrees of customization that might be available
    between "full functionality" and "one button" operation.

There are certainly other potential reference designs that could
                be used to explore those issues:
  * Phone plus Personal Digital Assistant (PDA). Wireless phones
      are now being manufactured that incorporate a larger
        touchscreen with stylus. In addition to hardware
    accessibility and usability issues (see Cress and French,
    1994), the phone-plus-PDA also provides software that is
      important to people with cognitive disabilities. How
   accessible and usable is the calendar/reminder program? The
     address book? The expense tracker? The Web browser? The
      email program? How would useful configurations of this
    device vary between someone with learning disabilities and
    someone with mental retardation? Are there a few default
     customizations that could easily be downloaded from the
         manufacturer's or service provider's Web site?

    * Phone-plus-PDA combined with desktop computer. How can
   desktop-PDA integration make both systems easier to use for
              people with cognitive disabilities?

  * Web style sheets. Service providers sending Web content to
   restricted displays (televisions, PDAs) filter and reformat
    it to fit. The content can become quite streamlined. How
    does this relate to features of content design that help
    people with cognitive disabilities better navigate and use
    the Web? Does it make sense to have specific style sheets
      for various cognitive disabilities, just as there are
    several large print, high contrast schemes for people with
    low vision? And how might this work generalize to talking
    books and to instructional CDs? (See Hendrix and Birkmire,
      1998; Nielsen, 1996; World Wide Web Consortium, 1998.)

                   6.2. SPECIFIC DESIGN ISSUES

            6.2.1. Layered or Restricted Functionality

     Between full functionality and one-button operation lie
  intermediate levels of cognitive difficulty. In the training
literature this is called a "training wheels" interface (Carroll
    and Carrithers, 1984; in Charness and Bosman, 1992). It is
 possible to measure how effective these interfaces are, both in
   and of themselves, and as a way to block typical errors and
    promote learning of the full product (Egan, 1988). In mass
market products, layered functionality shows up as "short menus"
       and feature-hiding panels. How effective have those
  implementations been, particularly for people with cognitive
  disabilities? Are there principled methods for determining how
 to layer or prune functionality? How might this be demonstrated
      in the effective design of a phone-plus-PDA device?

    6.2.2. Training, Procedural Support, and Problem-solving

  The adaptive training and tutoring systems described in this
  report (Anderson et al., 1995; Glisky, Schacter, and Tulving,
    1986; Glisky, 1992) were effective but time-consuming to
  develop. Yet mass market computer interfaces now have wizards
 and intelligent help systems that watch the user's behavior and
offer support. How well do these interfaces work for people with
  cognitive disabilities? How can they be improved? How can they
        be brought into the world of small but networked
                   telecommunications devices?

   In the area of problem-solving, we have the example of PEAT
  (Levinson, 1997), a problem-solving device that specializes in
  planning and scheduling. There are other problem-solving aids
 that could be redesigned for people with cognitive disabilities
and adapted to portable telecommunications devices. For example,
     people with cognitive disabilities need orientation and
    wayfinding (Vanderheiden, 1998b). One can begin to see the
  components of a better portable tool. First, on the Web there
are free mapping and route-planning services, but the directions
    may be difficult to follow for people with sequencing and
visualization impairments. Second, the "Atlas Speaks" product by
 Arkenstone takes pre-planned route information and downloads it
    into a portable device, in text or audio format. Finally,
  wireless phones will provide positioning information. If the
   positioning information is sufficiently accurate, it may be
possible for users not only to play back pre-planned directions,
 but also to find out where they are, determine whether they are
  off track, and request new directions from a network service.
 Such a device could also receive precise local information from
      Talking Signsr (Crandall, Bentzen, and Myers, 1998) in
            environments where those are available.

                6.3. USER PROFILES AND QUICK TESTS

  Designers have less understand of cognitive disabilities than
  sensory or mobility impairments. In our experience, designers
without access engineering background tend to think of cognitive
  disabilities as something like mental retardation, a broad and
 general intellectual impairment. They do not yet understand the
 specific deficits resulting from aging, learning disability, or
 brain injury. Nor, more importantly, do they yet understand the
  mapping between those specific deficits and equally specific
                        design solutions.

  Designers need more concrete pictures of people with cognitive
disabilities. These pictures, or "user profiles," help designers
      create usage scenarios for a product (Nielsen, 1993;
    Shneiderman, 1992). For example: "Jane, who has a specific
    learning disability, gets a reminder from her electronic
  appointment calendar and then sets up a conference call." User
  profiles and usage scenarios help designers before systematic
    user testing is possible, in the early stages of product
 definition. They are refined throughout the project. They serve
 as test cases: "Would this feature we are thinking about really
                           help Jane?"

  When these user profiles are developed, people with cognitive
  disabilities should be recruited as active participants in the
                  process (Doe and Whyte, 1995).

    It might even be possible to create "quick tests," rough
  simulations of what it is like to use products when one has a
  given cognitive disability. The goal of a quick test is not to
    be a precise simulation of human performance but to give
  designers a sense of the user experience. They are useful when
        formal testing with real users is not possible.


    Usability testing is only one method of evaluating human
  performance with a product. Human factors engineers can create
   user models, which are predictive simulations of a person's
                  performance with an interface.

  How long does it take to read the display? To decide which key
 to press? To press it? To evaluate the result? If one interface
     requires more keypresses, but another requires an extra
      decision, which is the better design - and for whom?

     User models are based on estimates of human perception,
cognitive processes, and motor behavior. These estimates can and
should be extended to include people with disabilities. Charness
  and Bosman (1990) combined the well-regarded methods of Card,
  Moran, and Newell (1983) with performance data from the aging
      literature to arrive at parameters for predicting the
    performance of 57.older adults. It is important to include
      adults over 75, since their cognitive capabilities are
  distinctly different from those of the "young-old" between 65
      and 75 (all citations in Charness and Bosman, 1992).

Given the time and effort that user modeling currently requires,
  it may be na=8Bve to think it will significantly reduce the need
for usability testing (as suggested by Dillon and Watson, 1996).
  However, the field of usability engineering needs design tools
  that cover people with disabilities as well as those without.


  Throughout this report we have highlighted research that ties
    specific design issues to equally specific human cognitive
capabilities. This correlative research (exemplified by the work
  of Egan, Glisky, Morrow, and various researchers on cognitive
                aging) has two important benefits.

    First, it helps to refine existing design strategies. For
  instance, because of the work of Morrow and colleagues on icon
 design (1998b), designers know more about implementing the TAAC
    strategy of including illustrations in documents (1999).

    Second, correlative research is a principled way of tying
    accessible design to universal design. It recognizes the
  underlying cognitive ability (e.g., working memory). It also
      acknowledges the wide variation in that ability within
        populations as well as between them (Egan, 1988).

  In accordance with the EITAAC approach (1999), priority should
  be given to research on closing the gap in performance between
 people with disabilities and those without. This will also help
      equalize performance within nondisabled populations.

  It's also important to include participants with disabilities
    who are less educated, wealthy, and independent than those
  typically recruited for studies. A rare example is the work of
   Craik, Byrd, & Swanson (1987, in Park, 1992). They compared
  recall performance of low-income, inactive elderly people with
 that of low-income active elderly, upper income active elderly,
  and young active adults. More structured recall cues helped to
  close the gap: the most disadvantaged, worst performing group
  made the largest improvements in recall. More attention also
    needs to be paid to older aging populations, those over 75
    (Charness and Bosman, 1992; Echt, Morrell, and Park, 1998;
              Science Directorate of the APA, 1993).

 The design topics of this correlative research will, of course,
    change as technology changes. The text editor improvements
    proposed by Egan and Gomez (1985) are commonplace today.
 Researchers in this area need to choose design topics that have
the greatest possible impact given technology trends and product

                        6.6. THE ROADMAP

 Finally, there is the roadmap discussed in section 1.3. Without
  this unifying framework, it is difficult to be clear about the
   relationships between these four factors in the literature:
    * Distinct groups of people with cognitive disabilities.

  * Real-world tasks or functions with which these various user
                    groups have difficulty.

  * Underlying cognitive factors creating those difficulties.

    Specific design changes that compensate for the cognitive
    deficits and improve task performance. Correlative studies
 (section 6.5) and reference designs (section 6.1) contribute to
    the roadmap, but need to be tied to a better theoretical
  framework (such as Carroll, 1993). User profiles (section 6.3)
 and underlying demographics are also a part of the roadmap, and
   are useful to marketers who need to understand the business
             implications of serving these markets.


                     APPENDIX 1 : REFERENCES

 Allen, Bryce. 1994. "Cognitive abilities and information system
   usability." Information Processing and Management , 30 (2),

  Alliance for Technology Access. 1996. Second Edition. Computer
  Resources for People with Disabilities: A Guide to Exploring
Today's Assistive Technology . Hunter House Publishers, Alameda,

  American Association on Mental Retardation. 1999. "Fact sheet:
                  What is mental retardation?"
  Web: http://www.aamr.org/Policies/faqmentalretardation.html.

      Anderson, John R., Corbett, A.T., Koedinger, K.R. & R.
Pelletier. 1995. "Cognitive tutors: Lessons learned." Journal of
              the Learning Sciences , 4(2), 167-207.

  Barker, Peggy. 1998. "Access Strategies for WebTV (R)." Paper
      presented at 1998 CSUN Conference, Northridge, CA..
         Web: http://www.dinf.org/csun_98/csun98_036.htm

     Bergman, Marilyn M. 1991a. "The necessity of a clinical
  perspective in the design of computer prostheses." Journal of
            Head Trauma Rehabilitation6 (2): 100-104.

 Bergman, Marilyn M. 1991b. "Computer-Enhanced Self-Sufficiency:
   Part 1. Creation and Implementation of a Text Writer for an
  Individual with Traumatic Brain Injury." Neuropsychology5 (1):

        Bergman, Marilyn M., and A. Gail Kemmerer. 1991.
"Computer-Enhanced Self-Sufficiency: Part 2. Uses and Subjective
Benefits of a Text Writer for an Individual with Traumatic Brain
            Injury." Neuropsychology , 5 (1), 25-28.

Bergman, Marilyn M. 1998. "Cognitive Orthotics Enhance the Lives
  of Users with Cognitive Deficits." Paper presented at the 1998
                CSUN Conference, Northridge, CA.
        Web: http://www.dinf.org/csun_98/csun98_160.htm

 Blanchard, Harry E. 1997. "Standards: Standards for Multimedia,
    Accessibility, and the Information Infrastructure." SIGCHI
                         Bulletin29 (3).
  Web: http://www.acm.org/sigchi/bulletin/1997.3/standards.html

     Brink, Satya. 1997. "The Twin Challenges of Information
  Technology and Population Aging." Generations: Journal of the
American Society on Aging: Aging and Information Technology , 21
                            (3): 7-10.
         Web: http://www.dinf.org/csun_98/csun98_160.htm

   Broderbund Software and The Alliance for Technology Access,
    1997. "An Evaluation of Broderbund Software Products: for
    children with learning disabilities and distinct learning
                styles." Broderbund Software, Inc.

  Brodin, Jane. 1992. "Telecommunication for People with Mental
  Retardation. Requirements and Services." Report No. 3 in the
  series Technology , Communication and Disability . Stockholm:
          Department of Education, Stockholm University.

    Brodin, Jane. 1994. "Videotelephony for two persons with
      moderate mental retardation." International Journal of
              Rehabilitation Research 17: 357-363.

 Brodin, Jane, and Magnus Magnusson. 1992a. "Videotelephones and
  mental retardation: survey of results achieved and research in
 progress." Report No. 1 in the series Technology, Communication
 and Disability . Department of Education, Stockholm University:

    Brodin, Jane, and Magnus Magnusson. 1992b. "Still Picture
    Telephones for People with Aphasia and Mental Retardation.
    Report No. 4 in the series." Technology, Communication and
   Disability . Department of Education, Stockholm University:

  Carroll, John B. 1993. Human cognitive abilities: A survey of
  factor-analytic studies. New York: Cambridge University Press.
                        622-627, 631-655.

  Charness, Neil, and Elizabeth A. Bosman. 1992. "Human factors
    and age." In F.I.M. Craik and T.A. Salthouse (Eds.), The
  Handbook of Aging and Cognition . New Jersey: Lawrence Erlbaum
                      Associates. 495 - 549.

Chute, Douglas L., Conn, Gretchen, Dispasquale, Madeline C., and
  Melanie Hoag. 1988. "ProsthesisWare: A New Class of Software
Supporting the Activities of Daily Living." Neuropsychology . 2:

    Cole, Elliot, and Parto Dehdashti. 1998. "Computer-based
cognitive prosthetics: Assistive technology for the treatment of
  cognitive disabilities." Presented at ASSETS '98: ACM SIGCAPH
 Conference on Assistive Technologies. Marina del Rey, CA, April

  Conder, Robert L. 1992. "Computer Technology in Brain Injury."
                   Brain Injury. 6 (1): 79-80.

      Cramer, Shirley C., and William Ellis. 1996. Learning
     Disabilities: Lifelong Issues . Baltimore: Paul Brookes
                      Publishing. xxvii-xxx.

  Crandall, Bill, Bentzen, B.L., and Linda Myers. 1998. "Talking
 Signs (R) Remote Infrared Audible Signage for Transit Stations,
  Surface Transit, Intersections and ATMs." Paper presented at
              1998 CSUN Conference, Northridge, CA.
         Web: http://www.dinf.org/csun_98/csun98_063.htm

    Cress, Cynthia J., and Clifford C. Goltz, 1989. "Cognitive
   factors affecting accessibility of computers and electronic
             devices." RESNA 12th Annual Conference.

  Cress, Cynthia J., and JoAnn P. Tew. 1990. "Cognitive skills
  associated with the operation of various computer interfaces."
            RESNA 13th Annual Conference, 251 - 252.

  Cress, Cynthia J., French, Greta J., and JoAnn P. Tew. 1991.
    "Age-related difference in interface control in normally
  developing children." RESNA 14th Annual Conference, 257-259.

  Cress, Cynthia J., and Greta J. French. 1992. "A framework for
    characterizing AAC device learning for persons with mental
          retardation." RESNA International Conference.

  Cress, Cynthia J. and Greta J. French. 1994. "The relationship
  between cognitive load measurements and estimates of computer
     input control skills." Assistive Technology , 6, 54-66.

    Day, Sheryl L., and Barbara J. Edwards. 1996. "Assistive
       Technology for Postsecondary Students with Learning
Disabilities." Journal of Learning Disabilities 29 (5): 486-492.

    DeVendra, Dennis P. 1998. "Features and Benefits for the
  Interface of the Next Generation Digital Talking Book." Paper
        presented at 1998 CSUN Conference, Northridge, CA.
         Web: http://www.dinf.org/csun_98/csun98_115.htm

 Dillon, Andrew, and Charles Watson. 1996. "User analysis in HCI
 - The historical lessons from individual differences research."
  International Journal of Human-Computer Studies , 45, 619-637.

Doe, Tanis. & John Whyte, 1995. "Participatory Action Research."
   Paper presented at the National Institute on Disability and
   Rehabilitation Research national symposium on Participatory
                Action Research, Washington, D.C.

  Echt, K.W., Morrell, R.W., & D.C. Park. 1998. "Effects of age
   and training formats on basic computer skill acquisition in
        older adults." Educational Gerontology, 24, 3-25.

 Egan, Dennis E. 1988. "Individual differences in human-computer
  interaction." In M. Helander (Ed.), Handbook of Human-Computer
       Interaction . Elsevier Science Publishers, 543-568.

Egan, Dennis E., & Louis M. Gomez 1985. "Assaying, isolating and
    accommodating individual differences in learning a complex
skill." In R. Dillon (Ed.), Individual Differences in Cognition,
          Volume 2 . New York: Academic Press, 173-217.

    Electronic Industry Foundation 1996. "Resource Guide for
        Accessible Design of Consumer Electronics." Web:

Electronic and Information Technology Access Advisory Committee.
       1999. "EITAAC Report, May 13, 1999." (Report to the
    Architectural and Transportation Barriers Compliance Board
                        [Access Board]).
            Web: http://trace.wisc.edu/docs/eitaac/.

 Epilepsy Foundation. "Examples of Job Accommodations for People
             with Epilepsy." Web: http://www.efa.org

      Glisky, Elizabeth. 1992. "Acquisition and transfer of
     declarative and procedural knowledge by memory-impaired
  patients: A computer data-entry task." Neuropsychologia , 30
                          (10), 899-910.

   Glisky, Elizabeth L., Schacter, D.L. & Endel Tulving. 1986.
 "Computer learning by memory-impaired patients: Acquisition and
   retention of complex knowledge." Neuropsychologia , 14 (3),

  Goldberg, Larry. 1996. "Access for People with Disabilities."
  Position paper at the Computer Science and Telecommunications
          Board (CSTB) of the National Research Council.

Hammill, Donald, Leigh, James, McNutt, Gaye, and Stephen Larsen.
  1987. "A new definition of learning disabilities." Journal of
            Learning Disabilities , 20 (2), 109-113.

    Haugen, K. 1996. "Accessible E-Mail: Motivating Literacy."
        Center for Accessible Technology . Berkeley, CA.

  Haugen, K. 1997. "Internet On-Ramps for Kids ... Point the Way
  with an Overlay." Center for Accessible Technology . Berkeley,

    Hendrix, Paul, and Michael Birkmire. 1998. "Adapting web
    browsers for accessibility." Paper presented at 1998 CSUN
                   Conference, Northridge, CA.
         Web: http://www.dinf.org/csun_98/csun98_113.htm

Hubble Dahlquist, Lori. 1998. "Classroom Amplification: Not Just
 for the Hearing Impaired Anymore." Paper presented at 1998 CSUN
                   Conference, Northridge, CA.
         Web: http://www.dinf.org/csun_98/csun98_124.htm

   Human Capital Initiative Coordinating Committee. 1993. "The
    changing nature of work." American Psychological Society
            Observer , October (Special Issue), 1- 24.

 Hunt, Earl. 1995. "The role of intelligence in modern society."
                    American Scientist. Web:

Job Accommodation Network. undated a "Case Summaries." A service
    of the President's Committee on Employment of People with
                  Disabilities . Morgantown, WV.

 Job Accommodation Network. 1997b "Worksite Accommodations Ideas
for Persons with Epilepsy or Seizure Disorder." A service of the
 President's Committee on Employment of People with Disabilities
                        . Morgantown, WV.

 Job Accommodation Network. 1997c "Worksite Accommodations Ideas
 for Persons with LD and / or ADD." A service of the President's
      Committee on Employment of People with Disabilities .
                         Morgantown, WV.

  Job Accommodation Network. 1995d "Worksite accommodations for
      people with cognitive disabilities." A service of the
 President's Committee on Employment of People with Disabilities
                        . Morgantown, WV.

  Johnson, Doris, and Jane Blalock. (1987) Adults with Learning
  Disabilities: Clinical Studies. Orlando, FL: Grune & Stratton,
                      Inc. 35, 82, 154, 208.

  Kaplan, Deborah, John De Witt, Maud Steyaert. 1992. Laying the
 Foundation: A Report of the First Year of The Blue Ribbon Panel
    on National Telecommunications Policy. World Institute on
                    Disability, Oakland, CA.

   Kerscher, George & Kjell Hansson. 1998. "DAISY Consortium -
 Developing the Next generation of Digital Talking Books (DTB)."
 DAISY Consortium presented at 1998 CSUN Conference, Northridge,
        Web: http://www.dinf.org/csun_98/csun98_ 065.htm

 Kirsch, Ned L., Levine, Simon P., Lajiness- O'Neill, Renee, and
  Marjorie Schnyder. 1992. "Computer-assisted interactive task
guidance: Facilitating the performance of a simulated vocational
    task." Journal of Head Trauma Rehabilitation. 7 (3) 13-25.

  Kreutzer, Jeffrey, Conder, Robert, Wehman, Paul, and Catherine
      Morrison. 1989. "Compensatory Strategies for Enhancing
  Independent Living and Vocational Outcome Following Traumatic
      Brain Injury." Cognitive Rehabilitation. 7 (1) 30-35.

  LaPlante, Mitchell, and Dawn Carlson. 1996. "Disability in the
     United States: Prevalence and Causes, 1992." Disability
 Statistics Rehabilitation Research & Training Center, Institute
for Health & Aging. University of San Francisco, CA. Law, Chris,
     and Gregg Vanderheiden. 1998. "EZ Access Strategies for
  Cross-Disability Access to Kiosks, Telephones and VCRs." Paper
        presented at 1998 CSUN Conference, Northridge, CA.
         Web: http://www.dinf.org/csun_98/csun98_074.htm

 Levine, Simon P., Horstmann, Heidi M., and Ned L. Kirsch. 1992.
"Performance considerations for people with cognitive impairment
  in accessing assistive technologies." Journal of Head Trauma
                  Rehabilitation . 7 (3): 46-58.

 Levinson, Richard. 1995. "A Computer Model of Prefrontal Cortex
  Function." In Jordan Grafman, Keith J. Holyoak, and Fran=87ois
  Boller, eds. Structure and Functions of the Human Prefrontal
 Cortex. Annals of the New York Academy of Sciences, Volume 769.
      New York: The New York Academy of Sciences: 381- 88.

 Levinson, Richard. 1997. "The; Planning and Execution Assistant
 and Trainer (PEAT)." Journal of Head Trauma Rehabilitation . 12
    (2): 85-91. Lewis, Rena B. 1998. "Assistive Technology and
     Learning Disabilities: Today's Realities and Tomorrow's
Promises." Journal of Learning Disabilities . 31 (1): 16- 26, 54.

  MacArthur, Charles A. 1990. "Using Technology to Enhance the
    Writing Processes of Students with Learning Disabilities."
        Journal of Learning Disabilities. 29 (4): 344-345.

 Mann, William C., Hurren, Dianne, Charvat, Barbara, and Machiko
  Tomita. 1996. "The Use of Phones by Elders with Disabilities:
 Problems, Interventions, Costs." Assistive Technology : 8 (23):

      Miller, Brita and Jimmie Shreve. 1994. "Strategies and
  Accommodations" presentation handout at the 1994 LDA and Rebus
                Conferences, Northern California.

  Morrow, Daniel G., Carver, Lisa, Leirer, Von O., and Elizabeth
Decker Tanke. 1998a. "Age differences in comprehension of health
      information." Presented at the Seventh Cognitive Aging
                    Conference. April, 1998.

  Morrow, Daniel G., Hier, Catherine M., Menard, William E., and
  Von O. Leirer. 1998b. "Icons improve older and younger adults'
      comprehension of medication information." Journal of
    Gerontology: Psychological Sciences, 53B (4), 240 - 254.

   Morrow, Daniel G., and Von O. Leirer (in press). "Designing
    medication instructions for older adults." In D. Park, R.
 Morrell, & K. Shifren (Eds.), Processing of Medical Information
    in Aging Patients. Mahwah, NJ: Lawrence Erlbaum, 249-265.

    Morrow, Daniel G., Leirer, Von O., Andrassy, Jill M., and
     Elizabeth Decker Tanke. 1995. "Designing health service
   appointment reminder messages for older adults." Journal of
            Clinical Geropsychology, 1 (4), 293-304.

Nielsen, Jakob. 1993. Usability engineering. Academic Press. San
  Diego, CA. Nielsen, Jakob. 1995. Multimedia and hypertext: The
          Internet and beyond . Boston: AP Professional.

     Nielsen, Jakob. 1996. "Accessible Design for Users with
            Disabilities." SunSoft. http://www.sun.com

    Nussbaum, Bruce, and Robert Neff. 1991. "I can't work this
          thing!" Business Week, April 29,1991: 58 - 66.

  Park, Denise C. 1992. "Applied cognitive aging research." In
  F.I.M. Craik and T.A. Salthouse (Eds.), The Handbook of Aging
and Cognition . New Jersey: Lawrence Erlbaum Associates. 449-491.

  Raskind, Marshall, H. and Eleanor L. Higgins. 1998. "Assistive
       Technology for Postsecondary Students with Learning
  Disabilities: An Overview." Journal of Learning Disabilities,
                          31, 1: 27-40.

Raskind, Marshall H., Eleanor L. Higgins, and Kenneth L. Herman.
  (1997) "Technology in the Workplace for Persons with Learning
  Disabilities: Views from the Inside." In Learning Disabilities
   and Employment. Edited by Paul J. Gerber and Dale S. Brown:

Raskind, Marshall, and Neil Scott. "Technology for postsecondary
            students with learning disabilities." 242.

  Raskind, Marshall, and Tobey Shaw. 1998. "Assistive technology
          for individuals with learning disabilities."
        Web: http://www.dinf.org/csun_99/session1002.html

    Reid, D. Kim, Hresko, Wayne, and H. Lee Swanson. 1991. A
  Cognitive Approach to Learning Disabilities. 27-29; 151-152.

   Riviere, Adrienne. 1996. "Assistive Technology: Meeting the
   Needs of Adults with Learning Disabilities." National Adult
   Literacy and Learning Disabilities Center. A Program of the
                National Institute for Literacy.

  Rogers, Wendy A., and Arthur D. Fisk. 1997. "Automatic teller
  machines: Design and training issues." Ergonomics in Design,5,

  Rose, David. 1998. "Report of the developer working group on
              universal design of curriculum." CAST.

  Science Directorate of the American Psychological Association.
 1993. "Vitality for life: Psychological research for productive
 aging. Research initiatives concerning behavioral solutions for
    the maintenance of health and vitality in late adulthood."

  Sein, Maung K., Olfmann, Lorne, Bostrom, Robert P. & Sidney A.
    Davis. 1993. "Visualization ability as a predictor of user
learning success." International Journal of Man-Machine Studies,
                          39, 599-620.

  Shaw, Stan F., Cullen, Joseph P., McGuire, Joan M., and Loring
    C. Brinckerhoff. 1995. "Operationalizing a definition of
  learning disabilities." Journal of Learning Disabilities, 28
                          (9), 586-597.

Shneiderman, Ben. 1992. Designing the user interface: Strategies
     for effective human-computer interaction . Reading, MA:
  Addison-Wesley Steele, Richard D., Weinrich, Michael, Maria K.
  Kleczewska, and Gloria S. Carlson. 1991. Computer-aided visual
   communication for severely impaired aphasics. Human-Machine
     Integration, 1991 Progress Report, Report 45: 111- 112.

Telecommunications Access Advisory Committee. 1999. "Telecom Act
Accessibility Guidelines (TAAG) and Strategies [Working Draft]."

    Telecommunications Access Advisory Committee. 1997. Final
   Report. Access to telecommunications equipment and customer
    premises equipment by individuals with disabilities. Web:
          http://www.access-board. gov/pubs/taacrpt.htm

  Trace Research and Development Center. (1996) "Problems with
        Web: http://trace.wisc.edu/docs/smartcards/atm.htm

  Trace Research and Development Center. (1996) "Problems with
 automated ticket and information services on public transport."
    Web: http://trace.wisc.edu/docs/smartcards/transport.htm

  Trace Research and Development Center. (1996) "Problems with
                    public telephones." Web:

    Trace Research and Development Center. 1999. "Trace Center
               reference designs for cell phones."
             Web: http://trace.wisc.edu/docs/phones/

  Trace Research and Development Center 1996. "What Problems Do
            People with Disabilities Have? and Why?"
Web: http://trace.wisc.edu/docs/software guidelines/software.htm

  Vanderheiden, G.C., & Scadden, L. 1988. Considerations in the
   design of computers and operating systems to increase their
    accessibility to persons with disabilities (Version 4.2).
        Madison, WI: Trace Research and Development Center
   Web: http://trace.wisc.edu/docs/considerations/consider.txt

 Vanderheiden, Gregg. 1992. "A brief look at technology & mental
retardation in the 21st century." Mental Retardation in the Year
       2000 . Louis Rowitz, ed. New York: Springer-Verlag.

    Vanderheiden, Gregg C. 1994. Application software design
guidelines: increasing the accessibility of application software
to people with disabilities and older users . Madison, WI: Trace
Research And Development Center, University of Wisconsin-Madison.

  Vanderheiden, Gregg. 1998. "Cognitive and language impairments
  and their implications." Accessible Design. Madison, WI: Trace
Research And Development Center, University of Wisconsin-Madison.

 Vanderheiden, Gregg. 1998. " Assistive techniques & devices for
  persons with cognitive and language impairments." Accessible
   Design. Madison, WI: Trace Research And Development Center,
                University of Wisconsin-Madison.

  Vanderheiden, Gregg C. 1997 "Cross Disability Access to Touch
 Screen Kiosks and ATMs." Advances in Human Factors & Ergonomics
                          21A: 417 - 420

    Vanderheiden, Gregg C., Law, Chris, and David Kelso. 1998.
    "Cross-Product, Cross-Disability Interface Extensions: EZ
    Access." Paper submitted for the RESNA Annual Conference.

  Vanderheiden, Gregg C., Mendenhall, John, and Andersen, Tom.
  1993. Access Issues Related to Virtual Reality for People with
  Disabilities . Madison, WI: Trace R& D Center, University of

Vanderheiden, Gregg C. and Katherine R. Vanderheiden. Accessible
  design of consumer products. Part II:Disabilities and specific
      barriers to accessibility . Madison: Trace R&D Center,
                University of Wisconsin-Madison.

   Vanderheiden, Gregg C. and Katherine R. Vanderheiden. 1991.
  Accessible design of consumer products. Part III : Guidelines
    for more accessible design. Madison, WI: Trace R&D Center,
                University of Wisconsin-Madison.

W3C. 1998. Web Accessibility Initiative (WAI). W3C working draft
      of WAI Accessibility Guidelines: Page Authoring Web:

 Ward, Carolyn. 1990. "Design for All: Consumer Needs Assessment
    Project Year 2: Results of the Second Year of a Five Year
    Study." Washington, DC: Electronic Industries Foundation,
                Rehabilitation Engineering Center.

  West, Robin. 1989. "Planning practical memory training for the
    aged." In L.W. Poon, D.C. Rubin, and B.A. Wilson (Eds.),
  Everyday Cognition in Adulthood and Late Life. city: Cambridge
                    University Press. 573-597.

    World Health Organization. International Classification of
 Impairments, Disabilities, and Handicaps (ICIDH). Geneva, 1980.

Yesavage, Jerome A., Lapp, Danielle, and Javaid I. Sheikh. 1989.
  "Mnemonics as modified for use by the elderly." In L.W. Poon,
    D.C. Rubin, and B.A. Wilson (Eds.), Everyday Cognition in
   Adulthood and Late Life . city: Cambridge University Press.

Young, Stuart W. 1994. Event Control. Investigative Radiology 29
                        (5): 590-593. 70



There are 111 documents in the bibliography. The documents can be
grouped as follows:  n 23 discuss characteristics of people with
12 disabilities or more general human cognitive capabilities.  n
55 cover products currently used by people with cognitive
disabilities or evaluate particular interfaces as a function of
cognitive ability.  n 5 hypothesize uses of certain technologies.

28 contain design guidelines (including cross- disability
guidelines) or other organizing frameworks.  More information was
found on technology for people with learning disabilities than
for people with aging- related cognitive disabilities. Technology
for people with traumatic brain injury or mental retardation is
even less frequently documented. Also, there is significantly
more information about computers than about telephones or other
telecommunications equipment.

[Rest of bibliography omitted because of size -- over 400K]

End of Document
Received on Friday, 12 May 2000 01:09:43 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 16 January 2018 15:33:32 UTC