[Archive copy mirrored from: http://www.w3.org/Journal/5/s3.paciello.html]
[PREVIOUS]|[NEXT]|[TOC]
People with Disabilities Can't Access the Web!
Mike Paciello
Abstract
Information access for people with disabilities is creating numerous opportunities and challenges in the Information Highway community. In addition, as a result of the increasing paradigm shift by the publishing industry toward Internet and WWW-based document delivery systems, the importance of producing accessible information using electronic document mechanisms has increased immeasurably.
This paper is focused on the production of electronic documents for people with disabilities. However, the key principals involved in the design, production, and delivery of information apply regardless of the document medium. The paper will attempt to identify major problems in information and software design that deny access; cite successful products that can be used by people with disabilities to access publications; and point to resources that assist developers in creating accessible products in the future.
Introduction
The publishing paradigm shift is nearly complete: we have moved from a paper-based, typewriter- generated, hand-edited, printing press produced publication to a paperless, intelligent, WYSIWYG, software generated, WWW-published hyperdocument. Not only has the shift changed the way information is produced, it has changed the way individuals read that information.
For temporarily able bodied persons (TABs), the shift has resulted in increased availability to a global information set never before achieved. Because of this apparent increased availability of information the publishing industry has wrongly assumed that "what is good for the goose, is good for the gander."
The sad truth is that the proliferation of information does not guarantee its accessibility. Availability does not equate to accessibility. Where people with disabilities (particularly those with print disabilities) are concerned, thoughtless barriers to information are being constructed by electronic publishers. The barrier factor is increased by the magnitude of inexperienced online businesses and organizations who have correctly assessed the inexpensive cost of delivering information on the Internet but have inaccurately assumed that because it's on the Web, it must be easy to read--or access.
What are the key issues involving information accessibility for people with disabilities? Do solutions exist to assist publishers in the design, production, and delivery of accessible publications? Can publishers increase the accessibility and availability of their documents without sacrificing additional time, creativity, quality, personnel, and money?
The answers to these questions and available resources are discussed in the sections that follow.
Access Systems for People
with Disabilities
Generally speaking, people with disabilities require assistive or adaptive devices to assist them in rendering or viewing a document. Those in the disability technology field refer to these as "access systems." For example:
- The blind may require a document output of braille or synthetic voice.
- Those with low vision or dyslexia may need large text or spatial adjustments.
- Individuals who are deaf or the hearing impaired may require visual cues for electronic documents that include sound or audio events.
- The physically or mobility challenged may require the ability to use the document viewer or browser without being able to use a keyboard, mouse, or input device that requires a part of their body other than just their eyes or mouth.
Access systems for people with disabilities include (but are not limited to): screen magnifiers, refreshable braille displays, screen readers, synthetic speech, caption-ready monitors, or alternative keyboards. The important point here is that the designer of the information need not worry about producing several versions of specialized documents. Rather, the focus should be on designing the source document with a rich set of characteristics that can be subsequently rendered or viewed by a wider audience.
To emphasize, this is not a new technology. Those involved in online publishing know that a source document can be coded using, for example, symbolic reference tags that are recognized by the document processor and then rendered to plain text, PostScript, or browser-compatible output. The same pre- and post-processing capabilities can be refined to produce braille, large text, and synthetic-voice output documents. No doubt, with the advent of publishing mechanisms like HyTime and Digital Audio, natural language voice documents (NLVDs) are possible. Therefore, access is not only achieved for people with disabilities, but language barriers are also diminished. Indeed with W3C and SGML consortium support for style sheets and link process definitions, the ability to produce accessible information for all people, disabled or not, has never been greater. This is the essence of universal design.
Identifying Key Issues in
Information Accessibility
To identify the key issues related to information accessibility it is necessary to break down the publishing process into its natural phases: design, production, and delivery.
Design
In the design phase, the most common issues related to document inaccessibility involve the following:
- Complex notation
- Image rendering
- Multimedia features
- Navigation
Complex notation, including math and science, is extremely difficult to render in an acceptable format, especially for the blind. Most blind persons require an alternative format of the information, either ASCII text, Grade II Braille, or synthetic speech. Therefore, the proper rendition of the notation is critical to the reader. However, the challenge of rendering complex notation that tends to be graphical often requires a transformation process in which few information publishers are interested in investing.
Image rendering provides similar problems. Because the blind or low vision user is likely to be using alternative output (access systems), certain design considerations should be implemented--for example, using meaningful descriptive text in conjunction with figures, images, or other graphical entities within a document. Descriptive or alternative text has become a standard for Web-based documents. In fact, some authoring tools like HotMeTaL Pro, HotDog Pro, and Corel Web.Designer have built-in prompts for alternate text. Ideally, browser providers should then be able to properly display alternative text. The best implementation of alternate text display (are you listening Bill?) is Microsoft's Internet Explorer, which automatically displays alternative text using bubble-help type notes as the user passes over the image with the mouse cursor. What will it take to get Netscape into the picture?
Electronic documents that contain multimedia features, including sound or video clips, require additional attention. Keep in mind that anything that emits sound cannot be heard by the deaf. It may not be heard by the hard of hearing or, for that matter, anyone viewing the document in a noisy environment. If you believe this to be an impractical example, consider the current industry move to WebTV and public kiosks. While a specific solution for providing Web captioning does not exist today, this should be considered an important feature of the document. Logic dictates that if consumer electronic manufacturers can produce televisions and monitors that support internal captioning controls, browser manufacturers and server protocol developers can design the means for delivering captioning through browsers.
Descriptive video provides a blind or low vision user with additional narrative that is useful, sometimes critical, to their comprehension of an electronic document. The process simply requires the interjection of descriptive narration during the spots within the video that are not otherwise filled with sound effects or dialog. As a result, the blind or visually impaired viewer achieves increased comprehension of the video event.
The National Center for Accessible Media (NCAM) in Boston, Massachusetts, currently provides a service that implements descriptive video for the motion picture industry. Descriptive video and captioning are perfect examples of how the power of markup should be used to enhance the richness and accessibility of a document. NCAM have recently been awarded grants to assist them in the research, design, and delivery of Web-based information for public television.
Navigating an electronic document, particularly a hypertext document, is a challenge for anyone. Keeping track of where you've been, where you want to go, and then getting there can be a cybernightmare. Still, being able to visually navigate through a document has obvious advantages the blind or low vision user cannot easily imitate. A navigational cue as simple as providing colored text provides meaning and definition that the non-visual user cannot see. Therefore, there is a need to design solutions that implement audio cues in concert with visual cues.
Remember too that navigation is often closely tied with memory and consistent design. People with cognitive limitations simply require visual memory aids and simplified page design. An example of this can be found at the WebABLE! Web site (http://www.yuri.org/webable/). The designer, Colin Moock, implemented a system of visual cues consisting of opened and closed doors. The concept is basic to most people and is simple to learn and remember.
Navigational difficulties clearly present challenges to every user. Consider the difficulty a visual user has today and then imagine doing the same with your computer monitor turned off! Or try navigating through a multicolumn table or an online newspaper that contains multiple columns on a single page.
Without a doubt, navigation requires acute sensory awareness. Navigation is not just a document road map; it is not a linear link. Rather, good navigational design includes a combination of seeing, hearing, and "feeling" your way to a specific destination in a comprehensible way.
Design guidelines for Web pages
Creating Web pages has never been easier. Authoring tools have removed much of the associated complexity of the language. Additionally, where markup languages are concerned, HTML is about as easy as they come. Therein lies the dichotomy: while creating pages is simple, designing pages is not--particularly, designing accessible Web pages. Following is a brief list of guidelines for designing accessible Web pages:
- Use short, functional, descriptive text for text links within a page. Avoid using one- to two- word text links. Using functional text descriptions provides better navigation and coherent feedback to blind users who rely on synthetic speech to render a page.
- Always provide text descriptions for images. This guideline benefits visually impaired users, people who use text-based browsers, and individuals who turn off the ability to display inline graphics within GUI-based browsers. To accomplish this, use the ALT="text" attribute. The negative effect of not including alternative text is highlighted in Figure 1, which shows the Windham Hill Record company Web page.
In Figure 1, the seven image icons that border the left side of the page are menu-related icons that provide the user with navigational aids throughout the rest of the Web site. Note the use of ALT text for some images and not for others. How will a blind person comprehend or navigate this site?
- ALT text should be included for bulleted lists, horizontal rules, or thumbnail links to larger images.
Where images require lengthy text descriptions in order to convey proper definition, some Web masters have created a text anchor to a page. In this case, the preferred method has been to use the uppercase letter "D" to signify "descriptive
text." Users then click on a text anchor, and a full text version of the page is rendered. The WGBH National Center for Accessible Media accomplishes this nicely as noted in the examples in Figure 2.
Figure 3 shows the text anchor page that results once the user clicks on the "D" text anchor link.
- Consider using a text anchor page for tables. Synthetic voice software is capable only of reading left to right, one line at a time. The use of complex table constructs are very difficult for the blind reader to interpret.
- Use text anchors for image maps as logical connections to their associated links. This generally is done by placing the links directly below the image map. Designers may also consider providing users with a choice between viewing the image map or a text rendition (anchor page).
- Provide text transcriptions of audio clips or video clips that contain audio. This benefits deaf and hearing impaired users. Again, text anchors are the preferred method.
- Test your pages for color contrasts. Samu Mielonen recommends that you test your pages by adjusting your monitor to 256 shades of gray and then look for the following:
- Can you distinguish the different colors on the basis of their lightness values?
- Can somebody else who has not designed your screen discern the differences?
- Is the text pleasant to read or does it tire you?
- Are you using too many different colors (lightness values) that just confuse the reader without adding information to your design?
- Check for blue/yellow and red/green combinations--these color pairs are difficult or impossible to distinguish for those who have dichromatism (a form of colorblindness in which only two of the three fundamental colors can be distinguished due to a lack of one of the cone pigments).
- Avoid using frames. Frames cannot be easily read by the blind or visually impaired. Jakob Nielson also notes, "Splitting a page into frames is very confusing for users since frames break the fundamental user model of the Web page. All of a sudden, you cannot bookmark the current page and return to it (the bookmark points to another version of the frameset), URLs stop working, and printouts become difficult. Even worse, the predictability of user actions goes out the door: who knows what information will appear where when you click on a link?" [4][1]
- Do not use blinking text, scrolling text, marquees, or continuous animation.
- Provide forms that can be downloaded, then mailed, or emailed. Forms cannot be easily processed by the blind and visually impaired.
- Do not use browser-specific tags.
- Always test your pages with a variety of browsers.
Document Production
The accessibility issues related to the production of a document primarily involve the following:
- Authoring tools that are themselves accessible and also enforce accessible design tags, semantics, and protocols.
- The ability of the processor to receive a single source document and build accessible or alternative outputs.
For example, the IBM Bookmanager can build documents for the blind to use with their screen readers and voice synthesizers. This is because IBM Bookmanager supports the ICADD (International Committee for Accessible Document Design) DTD, which was designed to produce accessible documents for the print impaired. Documents produced for the World Wide Web are gradually becoming more accessible and require less "massaging" by a post processor or other intermediary actions because some Web browsers contain access features that enhance the readability of a document to persons with disabilities.
Document Delivery
In this case, document delivery refers to the ability of an online viewer or browser to adequately display a document. Most browsers or viewers were created for temporarily able bodied persons. They do not contain accessible features or controls that make it easier for people with disabilities to use. Additionally, they are rarely designed to allow assistive product manufacturers to easily link their products to using software "hooks."
A classic example is pwWebSpeak (The Productivity Works). This GUI browser was designed with synthetic speech and large text functionality built in. It also supports the HTML 2.0 specification, which includes the ICADD SGML Document Access attributes. The next section includes more information about pwWebSpeak.
Ensuring electronic document delivery through a browser could be significantly enhanced if developers included assistive preference options that allow a user to "turn on" captioning, descriptive video, sound cues, synthetic voice, keyboard mapping, screen magnification, and other accessibility features.
As noted earlier in this paper, people with disabilities are not completely unable to read or view electronic documents. Several solutions exist, which were designed to increase accessibility to information. The following sections provide manufacturer-supplied descriptions of each solution.
Successful Solutions
pwWebspeak Browser
The pwWebSpeak browser (The Productivity Works) is an Internet browser designed for users who wish to access the Internet in a non-visual manner. This includes users who cannot be tied to a keyboard or monitor, blind or visually impaired users, users with dyslexia or other learning disorders, and users who are learning new languages.
The intelligence built into pwWebSpeak understands the HTML constructs and automatically bypasses those constructs that have no relation to the information content of a document. Both speech and large character interpretation of the Web pages are provided so that all classes or users can use the software effectively.
pwWebSpeak is designed specifically to interact directly with the information on the Web pages and to translate the information content into speech. The user may navigate through the structure of a document based on its contents, paragraphs and sentences, rather than having to deal with scrolling and interpreting a structured screen display.
"Bobby" Accessibility Verifier
Bobby is a program that finds common accessibility problems on Web sites. It was created to help Web designers insure the greatest possible audience for their Web sites and especially to highlight common accessibility problems for those with disabilities. When Bobby analyzes a page, it provides detailed information on:
- Problem HTML tags and constructs that make it difficult for people with a variety of disabilities to access a Web page.
- Image and document load times to ensure quick page access on slow modems.
- A comprehensive analysis of how the page's HTML might be incompatible with those versions of HTML implemented by the major Web browsers (including Netscape Navigator, Internet Explorer, Mosaic, and AOL).
- Common problems with image maps.
- A variety of other important tests including looking for missing ALT text in image tags.
HTML-to-Braille Transformation Service
Both UCLA (http://www.ucla.edu/ICADD/html2icadd-form.html) and the University at Dresden (http://elvis.inf.tu-dresden.de/html2brl/) provide an HTML-to-braille transformation service. These services allow anyone who has an HTML-coded document to send that document to the server which then:
- Converts the document to the ICADD DTD, which may then be easily converted to braille, large text, or voice ready file format.
- Converts the document to the ICADD DTD and then converts that instance to a braille formatted file. This resulting file may then be sent to a braille printer or refreshable braille display for the blind user.
University of Toronto
SoftQuad together with the University of Toronto are presently engaged in a project to make SGML and HTML authoring and browsing tools accessible to people with disabilities and to guide SGML and HTML authors in creating accessible documents. The user who is reading hyperlinked multimedia documents using access technology such as screen readers, braille displays, or screen magnifiers, faces three challenges:
- Obtaining an overview of the document.
- Moving to specific sections or elements in the document.
- Retrieving alternative descriptions or labels of graphic or audio objects.
The user with sight can get a good sense of the content and scope of a document at a glance. Formatting and layout allow the user to quickly find a specific part of a document. This is not the case when using screen readers, braille displays, or screen magnifiers. SoftQuad and the University of Toronto are exploiting the structure inherent in SGML to allow users to efficiently navigate and obtain an overview of the document.
Available Resources
Several resources were relied on to develop this paper and are available to users, publishers, and software product manufacturers interested in information accessibility for people with disabilities. Following is a brief list of resources, including URLs.
Environment Canada's Adaptive
Computer Technology Centre: Accessible Web Page Design
Environment Canada's Adaptive Computer Technology Centre (http://www.igs.net/~starling/acces.htm) was one of the first Web sites to include guidelines and online examples of accessible Web pages and HTML implementations.
Adaptive Technology Resource Centre at the University of Toronto
The purpose of the ATRC (http://www.utoronto.ca/atrc/) is to:
- Develop and share creative solutions to the challenges faced by users of adaptive technology.
- Foster the effective use of adaptive technology in education.
- Promote the integration of alternative access systems throughout the information technology infrastructure at this and other educational institutions.
Recordings for the Blind and Dyslexic
Providing services for nearly half a century, Recording for the Blind & Dyslexic (RFB&D, http://www.rfbd.org) is a national nonprofit organization that serves people who cannot read standard print because of a visual, perceptual, or other physical disability. RFB&D is recognized as the nation's leading educational lending library of academic and professional textbooks on audio tape from elementary through post-graduate and professional levels.
Trace Research & Development Center's Designing an Accessible World
The Trace R&D Center (http://www.trace.wisc.edu/world/world.html) is one of the leading assistive technology research facilities in the world. Their Web page is a classic example of accessible design. Additionally, Trace provides several online reference documents.
WGBH National Center for
Accessible Media
NCAM (http://www.wgbh.org/ncam) develops strategies and technologies to make media accessible to millions of Americans, including people with disabilities, minority language users, and those with low literacy skills. For 85 million Americans with little or no access to media's sights and sounds, the CPB/WGBH National Center for Accessible Media (NCAM) is working to remove the barriers to communication by:
- Developing strategies and technologies to make all public media accessible at home, work, school, and in the community.
- Retrofitting existing television, radio, print, and theatrical movies with access technology.
- Designing access into emerging telecommunications such as Advanced Television (ATV) and the National Information Infrastructure.
Yuri Rubinsky Insight Foundation
The Yuri Rubinsky Insight Foundation (http://www.yuri.org) is dedicated to bringing together workers from a broad spectrum of disciplines to stimulate research and development of technologies that will ensure equality of access to information of all kinds. The YRIF is dedicated to commemorating the genius of the late Yuri Rubinsky.
The YRIF is also the new home of the WebABLE! disabilities information repository (www.yuri.org/webable/).
Ingram, Ray. "Universal Accessibility--A Matter of Design," The Productivity Works, Inc., Princeton, New Jersey. http://www.prodworks.com/ua_9606.htm
Gunderson, Jon. "World Wide Web Browser Access Recommendations," Mosaic Accessibility Project, Usability Access Chair, University of Illinois at Urbana/Champaign. http://www.staff.uiuc.edu/~jongund/access- browsers.html
Gunderson, Jon. "World Wide Web Accessibility to People with Disabilities: A Usability Perspective," Mosaic Accessibility Project, Usability Access Chair, University of Illinois at Urbana/Champaign. http://www.staff.uiuc.edu/~jongund/access-overview.html
Nielsen, Jakob. "Top Ten Mistakes in Web Design," SunSoft, Inc.
Vanderheiden, Gregg C., Wendy A. Chisholm, Neal Ewers, "Design of HTML Pages to Increase Their Accessibility to Users With Disabilities: Strategies for Today and Tomorrow," Trace R&D Center, University of Wisconsin, Madison. http://www.trace.wisc.edu/TEXT/GUIDELNS/HTMLGIDE/htmlgide.html
About the Author
Mike Paciello
131 D.W. Highway #618
Nashua, NH 03060
paciello@ma.ultranet.com
Michael Paciello is Executive Director of the Yuri Rubinsky Insight Foundation. The Foundation is dedicated to stimulating research and development of technologies that will ensure equality of access to information for all people.
Mr. Paciello currently serves as Chairman of the Electronic Industries Association's Assistive Devices Division (EIA/ADD). Mr. Paciello also serves as the W3C's Disabilities Guest Editor, is co-founder of the International Committee for Accessible Document Design (ICADD), and creator of WebABLE!, one of the world's largest Web sites dedicated to people with disabilities.
[1] See also Jakob Nielsen's article, "Guidelines for Multimedia on the Web," in this issue.
[PREVIOUS]|[NEXT]|[TOC]
