Search   |   Back Issues   |   Author Index   |   Title Index   |   Contents



D-Lib Magazine
September 2004

Volume 10 Number 9

ISSN 1082-9873

Library Web Accessibility at Kentucky's 4-Year Degree Granting Colleges and Universities


Michael Providenti
Web Development Librarian
W. Frank Steely Library
Northern Kentucky University

Red Line



Web sites designed to be accessible to all, using the principles of Universal Design, are not just a good idea, they are mandated by law. While Web accessibility has become a well-established topic in library literature, reaching back to at least 1996, there seems to be a disconnect between recommendations for and implementation of accessible design [1]. Many articles on the topic of Web accessibility refer to Universal Design (UD), which seeks to produce "products and environments useable by all people, to their greatest extent possible, without the need for adaptation or specialized design" [2]. According to the principles of UD, this outcome should be accomplished with little or no additional expense. For Web designers, the nuts-and-bolts of implementing UD are defined by the World Wide Web Consortium's (W3C) Web Content Accessibility Guidelines (WCAG) and Section 508 of the Rehabilitation Act of 1973 that, when implemented completely, result in sites as accessible as user agents allow.

The following survey of academic library web sites in Kentucky cannot be construed as a microcosm for academia in general. Only 11 states, Kentucky among them, have addressed Web accessibility in state laws and statutes [3]. If anything, based on the existence of state laws, Kentucky should be among the forward guard on this issue.

Disability Statistics

According to the 2000 U.S. Census, Arkansas, Kentucky, Mississippi and West Virginia had the highest disability rates in the country. For those aged 21 to 64, 24 percent of the population, nearly one in four people, reported some form of disability [4]. Statistics collected by the U.S. Census Bureau and U.S. Department of Commerce report that 3.5 percent of the U.S. population has vision problems which, as used here, means "difficulty seeing the words and letters in ordinary newsprint even when wearing glasses or contact lenses if they usually wear them, in addition to the blind." The same survey reports that 3.0 percent "has difficulty using hands" [5] Given in simple percentages, the numbers seem small enough. However, when we consider that we are talking about 18,511,800 people in the United States when we refer to the 6.5 percent with these specific disabilities, it becomes clear that this is not a tiny few. Extrapolated for Kentucky's population in 2001, approximately 264,300 people would be affected by these two disabilities within the state. This is without taking into consideration that Kentucky reports disabilities at higher rates than other states [6].

Turning from the U.S. population in general to statistics collected about college and university freshmen, the numbers are lower than those of the general population. For this group 1.0 percent report vision problems while only 0.4 percent report orthopedic disabilities [7]. One may speculate as to why these percentages are much lower than those of the general population. One possibility is that disabilities are reported in greater number as age increases and college freshman are at the low end of the age range. Another possibility is that the disabled may not view colleges and universities as being disability-friendly enough, despite legislation and programs to accommodate their specific needs.

Recent surveys of Internet use by those with disabilities provide more data by which to understand this user group. While it is reported that 42.3 percent of the U.S. population in 2000 with no disabilities used the Internet, 15.1 percent with disabilities did [8]. A Harris Poll, also from 2000, sought data regarding the importance of Internet access to the disabled. The Harris Poll findings revealed that adults with disabilities spent twice as much time online as those without disabilities. In addition, those with disabilities reported in far greater numbers than those without disabilities that the Internet had greatly improved their quality of life and helped them become better informed about the world [9]. Assistive technology, such as screen readers which can convert online text to speech, can certainly make an online newspaper far more accessible than the print equivalent to those with vision problems. While online access has made some things potentially easier for those with disabilities, increased bandwidth capabilities has led to an increase in graphics which, in turn, has unnecessarily diminished user-friendliness for the disabled [10].

For those who consider these numbers too small to be concerned with, there are two things to keep in mind. First, creating a web site that works well with assistive technology, specifically by making a site that conforms to current Web scripting and accessibility standards as defined by the World Wide Web Consortium (W3C), will ensure forward-compatibility for people using new user agents such as wireless personal digital assistants (PDAs) and web-accessible cell phones [11]. Second, the need to provide accessibility is mandated by law.

Web Accessibility and the Law

Beginning with Section 504 of the Rehabilitation Act of 1973 (29 USC 794), it became illegal for entities receiving federal funds to discriminate on the basis of disability. Clearly, given the date of the Act, there is no reference given to the Internet or online accessibility, but neither is the language restricted as to exclude them [12]. Just on the basis of Section 504, Steve Noble, Policy Analyst for the Kentucky Assistive Technology Service Network, writes:

...libraries and other institutions receiving federal funds have a clear responsibility under Section 504 to ensure that people with disabilities have an equal, effective, and inclusive opportunity to participate in any Web-based service or computer-based application that it offers to others [13]

The language of Section 504 (29 USC 794 (a)) seems clear enough:

No otherwise qualified individual with a disability in the United States, as defined in section 706 (20) of this title, shall, solely by reason of his or her disability, be excluded from the participation in, be denied the benefits of, or be subjected to discrimination under any program or activity receiving Federal financial assistance.

The phrase "program or activity" is defined under Section 504 to include "a college, university, or other postsecondary institution, or a public system of higher education." In 1986 Section 508 of the Rehabilitation Act of 1973 was added to require federal agencies to use, whenever possible, accessible technologies [14].

The passage of the Americans with Disabilities Act of 1990 (ADA) expanded the Section 504 legislation to include private entities that do not receive federal funds. Under ADA Title II (which covers state and local governments) and Title III (which addresses the private sector), reasonable modifications and auxiliary aids are obligatory to ensure "effective communication," which, the U.S. Department of Justice (USDOJ) explained in a letter to Senator Tom Harkin, includes the Internet [15]. The standard under ADA Title III, section 36.10, is described as follows:

No individual shall be discriminated against on the basis of disability in the full and equal enjoyment of the goods, services, facilities, privileges, advantages, or accommodations of any place of public accommodation (emphases added).

The current point of contention in the courts is whether the ADA's "place of public accommodation" refers narrowly to the physical environment or the digital online environment as well [16]. According to an amicus brief filed by the USDOJ in Hooks v. OKBridge, the Internet is a place of public accommodation though this issue was not addressed in the settlement of the case [17]. It was determined that since OKBridge was not aware of the plaintiff's disability, the ADA had not been violated and the place of public accommodation was, therefore, not an issue [18]. Since private clubs are exempt under the ADA, it is possible that OKBridge may have used such status as a defense had the case not terminated [19]. In a paper presented for the Bar Association of San Francisco Barristers High Tech Law Committee, Aaron Silberman writes that "A site that provides access to a physical space that itself qualifies as a place of public accommodation is probably subject to disabled access law requirements" [20]. The reasoning is that the Web sites are "communications providing access to places that clearly are public accommodations" and that the ADA "requires places of public accommodation to 'ensure effective communication with individuals with disabilities' (28 C.F.R. §36.303 (c))" [21] In Doe v. Mutual of Omaha Insurance Company, the Seventh Circuit Court took the position that physical space was not a requirement of a place of public accommodation [22].

The Workforce Investment Act of 1998 amended Section 508 of the Rehabilitation Act of 1973 and required guidelines to be written defining accessibility from a regulatory perspective [23]. The 16 guidelines, labeled (a) to (p) and outlined in § 1194.22, were written by the Architectural and Transportation Barriers Compliance Board in 2000 and went into effect on February 20, 2001 [24]. The purpose of these guidelines was to ensure that:

Federal employees with disabilities have access to and use of information and data that is comparable to the access and use by Federal employees who are not individuals with disabilities, unless an undue burden would be imposed on the agency [25].

While Section 508 only explicitly applies to Federal agencies, the U.S. Department of Education has included Section 508 compliance as part of its regulations enforcing the Assistive Technology Act of 1998 (29 USC 3001) [26]. In short, state agencies that receive Assistive Technology Act funding must abide by the Assistive Technology Act, which requires compliance with the Technology-Related Assistance for Individuals with Disabilities Act of 1988 (PL 100-407), which, in turn, calls for compliance with Section 508 [27].

In a letter to Dr. James Rosser, President of California State University, the US Department of Education's Office for Civil Rights (OCR) has addressed the "undue burden" language that appears in Section 504, 508, and the ADA. The OCR letter states that:

the subsequent substantial expense of providing access is not generally regarded as an undue burden when such cost could have been significantly reduced by considering the issue of accessibility at the time of the initial selection [28].

While the OCR's letter refers to hardware and software specifically, this could be construed to imply that the time and cost involved in a Web site's accessibility retrofit or redesign would not constitute an undue burden since barriers to accessibility could have been avoided in the initial design.

In addition to the Rehabilitation Act of 1973, its amendments, and the ADA, the Legislature of the Commonwealth of Kentucky has enacted statutes requiring Web accessibility. The Kentucky Revised Statutes (KRS) indicate that any state or state-assisted organization shall provide access to individuals with disabilities in compliance with national standards such as Section 508 [29].

The language which defines the standard of access varies between Section 508, the ADA, and the KRS. The language in Section 508, quoted above, requires comparable access for those with disabilities and those without. The KRS, which references Section 508, requires equivalent access for those with disabilities and those without [30]. The ADA, under Title III, offers a potentially higher standard, also quoted above, "full and equal enjoyment of the goods, services, facilities, privileges, advantages, or accommodations." The specifics of "full and equal" under the ADA are defined in Title III sec. 36.202 (a)-(c) which does not allow "(a) Denial of participation," "(b) Participation in unequal benefit," or "(c) Separate benefit." It is possible that the following recommendation under Section 508 could be construed as a separate benefit:

(k) A text-only page, with equivalent information or functionality, shall be provided to make a web site comply with the provisions of this part, when compliance cannot be accomplished in any other way. The content of the text-only page shall be updated whenever the primary page changes.

In the event that the text-only version is not carefully maintained, the site would be in violation of both Section 508 and the ADA. While Title II of the ADA, which addresses public institutions, does not use the "full and equal" language of Title III, it does indicate in Section 202 that individuals with disabilities shall not be excluded, shall not be denied benefits, and shall not be subject to discrimination. In either case, barriers to Web accessibility, whether the result is described as denial of full and equal privileges, or described as exclusion, denial of benefits, and discrimination, are contrary to the ADA.

Even if the courts were to eventually conclude that the ADA did not apply to Internet sites and Section 508 only applies to federal agencies, librarians have an ethical responsibility to provide equal access under the Library Bill of Rights. Under the Library Bill of Rights:

Libraries which make exhibit spaces and meeting rooms available to the public they serve should make such facilities available on an equitable basis, regardless of the beliefs or affiliations of individuals or groups requesting their use [31].

It would be unlikely that the American Library Association would argue that "exhibit spaces" did not extend to the Internet.

Accessibility and Web Standards

The authors of the Section 508 guidelines recognized the W3C's Web Accessibility Initiative's (WAI) Web Content Accessibility Guidelines (WCAG) 1.0 as a pre-existing standard [32]. The WAI first met in 1997 to begin work on the technical issues surrounding accessibility [33]. On May 5, 1999, the WAI issued the WCAG which gave web developers a concrete set of rules by which to construct accessible Web sites [34]. The WCAG are made up of 14 guidelines, each divided into various "checkpoints." Not to confuse the issue, three priority levels define the importance of each checkpoint: failing to satisfy Priority 1 issues will make accessing the site "impossible" for some; failing to satisfy Priority 2 issues will make accessing the site "difficult" for some; and failing to satisfy Priority 3 issues will make accessing the site "somewhat difficult" for some. Considering the "full and equal" language of Title III and the anti-discrimination language in Title II of the ADA, it would seem that nothing less than full compliance would suffice.

While the W3C was founded to create open interoperable standards, it cannot mandate the acceptance of these standards. In 1998, independent of the W3C, the Web Standards Project (WaSP) was founded to persuade software manufacturers, such as Netscape and Microsoft, to accept open standards as defined by the W3C [35]. Finally, with the release of Internet Explorer 6.0 in 2001, Netscape 6.0 in 2001, Mozilla 1.0 in 2002, and Opera 7.0 in 2002, Web sites could be built correctly to W3C standards [36].

What are Web standards? "Web standards" was a phrase coined by WaSP to replace the term "recommendation" used by the W3C. Web standards include Extensible Hypertext Markup Language (XHTML), Cascading Style Sheets (CSS), the WCAG, and a host of technologies developed by the W3C [37]. XHTML markup, when used correctly to standard, provides information about a Web page's content. This is important because assistive technology for the disabled reads markup. Page headlines need to be marked up as headlines using <h1>, long quotations need to be marked up as block quotes using <blockquote>, and lists need to be appropriately marked up as lists. Web pages that don't follow standards ignore the structural meaning of markup and focus on presentational effect. In non-standard Web pages the deprecated <font> tag is used to create the appearance of a headline, block quote tags to indent elements, and tables or line breaks to simulate lists. While this may result in a visually attractive page, markup used for non-structural semantic purposes makes pages confusing for those depending on screen readers.

Because XHTML structural markup alone results in a plain page, CSS is used to control presentation. Consider a navigation menu on a web page. Thinking of a menu in terms of structural markup would suggest an unnumbered list. CSS could be used to remove the traditional bullets marking list items and produce the appearance of buttons. In a browser that correctly renders CSS, the list items would have the appearance of graphics, while in old browsers without CSS support, the menu would appear logically as a simple bulleted list. By separating content marked up with XHTML and presentation controlled with CSS, the site can not only be redesigned by replacing CSS files but various CSS files can be written to allow users to select a presentation that best suits their needs. The key to making this work well is to structure pages logically so that the Web pages degrade gracefully in older equipment. One example of a logically structured page could start with a level 1 headline identifying the site, an option to skip to the page navigation, a level 2 headline identifying the page, the main content, the navigation, and then the footer including the address.

Embracing standards-based design requires a paradigm shift. Designers creating pixel-perfect Photoshop-drawn Web pages need to step back and refocus on user-centered issues. WCAG guideline 3.4, "Use relative rather than absolute units," addresses this. Relative units allow for users to resize fonts and for pages to flow comfortably in a wide range of monitor settings. Committees charged with oversight of Web design need to stop and read the WCAG and allow the standards to direct the process. Consider WCAG guideline 11.1 which recommends using the most recent supported W3C technology. Following this guideline would mean a new design should be built with XHTML 1.1 and CSS2. Guideline 3.2, the "validate to published formal grammars" recommendation, directs designers to validate their code, a process which tests the HTML and CSS for errors, in an online validator [38]. Valid code, unlike incorrect or proprietary code, is platform independent. Guideline 6.1, "Organize documents so they may be read without style sheets," requires logical document structure. Taking the whole of the WCAG into account at the beginning of a design project will certainly result in greater usability of the final product.

Recommending complete adherence to the WCAG is not meant to suggest the guidelines are perfect or complete. WCAG guideline 9.5, "Provide keyboard shortcuts to important links," while important, can be problematic if implemented according to W3C recommended techniques. The W3C demonstrates keyboard shortcuts, using "accesskey", defined with letters which can override some browser functions [39]. Shortcuts defined with numbers present fewer potential problems [40]. The point is that the WCAG and Section 508 should be understood before Web site building begins, should provide direction to the designer or committee, and should be implemented fully in the spirit, if not the letter, of the guidelines. The goal of standards implementation is maximum accessibility and maximum usability, not minimum compliance.

Selected Literature Review

The first study relevant to this current article was published by Erica B. Lilly and Connie Van Fleet in 1999, "Wired But Not Connected: Accessibility of Academic Library Home Pages" [41]. For this study, Yahoo!'s list of the "100 Most Wired Colleges" determined which institutions would be examined. Bobby 3.0, an accessibility tester then available from the Center for Applied Special Technology (CAST), was used to determine which sites were free from barriers to accessibility. Forty library home pages were determined to be accessible (i.e., they passed the equivalent of the Priority 1 automated checkpoints, meaning they did not lack alternative text equivalents, did not use server side image maps, and avoided blinking text). Statistical analysis showed no relationship between a university's resources or library's ranking and page accessibility. In an article titled "The Accessibility of Web Pages for Mid-Sized College and University Libraries" published in 2002, Tim Spindler used Bobby 3.2 software to examine the library Web sites at the 190 colleges and universities identified at as having between 5 and 10 thousand students [42]. Spindler found 42 percent of sites tested passed Priority 1 automated checkpoints, but concluded that the situation was not dire since half of the sites which failed to pass had fewer than 5 simple errors.

Axel Schmetzke has published, both on his personal Web site at the University of Wisconsin and in journal literature, a series of studies and articles on the accessibility of library Web sites. His well researched article "Web Accessibility at University Libraries and Library Schools" examined the relationship between the accessibility of Web sites of the top 24 library science programs as identified by US News and World Report and the accessibility of the library's Web site at the same institution [43]. Using a slightly newer version of Bobby than that used by Lilly and Van Fleet, Bobby 3.1.1, Schmetzke found 59 percent of library Web sites passed Priority 1 automated checkpoints compared with 23 percent at schools of library and information science. Statistical analysis was used to conclude that "There is a mild tendency for the Web pages of the more highly ranked schools of library and information science to be more accessible" (my emphasis) [44]. As for the connection between schools of library and information science and the libraries at the same institutions, no relationship was found to exist [45].

Schmetzke's annual surveys of Web accessibility at University of Wisconsin campuses, beginning in 1999, are a valuable source of longitudinal data generated at a single institution [46]. While data assembled from 1999 to 2002 only looked at Priority 1 accessibility issues, the 2003 survey was expanded to include all three Priorities as well as Section 508 compliance. His surveys show that, at least at the University of Wisconsin, accessibility was improving over time; a situation which may be a result of his surveys. His site also includes a long list of articles and surveys on the topic of Web accessibility and libraries.

On the topic of making library web sites more accessible, a short list of articles would include Sue Polanka and Jack O'Gorman's "Guidelines for Creating Accessible Library Web Pages," Sean Kinder's "Web Page Accessibility for Users with Disabilities," and Cheryl Kirkpatrick's "Getting Two for the Price of One: Accessibility and Usability" [47]. Each of these provides information on implementing basic accessibility guidelines defined in the WCAG. For a more in-depth discussion of the issues, we must look outside of library literature. For an understanding of how Web design impacts users with various disabilities, nothing approaches Mark Pilgrim's Dive into Accessibility, a free online book [48]. Pilgrim deftly and concisely describes problems encountered by people with specific disabilities and provides easy-to-follow solutions. Joe Clark's Building Accessible Websites does not simply provide recommendations; it explores accessibility barriers in-depth as well as the problems encountered implementing some solutions [49]. For an understanding of what we mean by Web standards and how to use them for accessibility, there is Jeffrey Zeldman's Designing with Web Standards [50]. These last three titles are essential reading for any designer or committee contemplating an accessibility retrofit or redesign.


The institutions in Kentucky to be examined were identified by the CollegeSource Online database. The advanced criteria search was used to search by state (Kentucky) and degree level (Bachelors). Thirty-three institutions were found and the location of their library home pages was identified by browsing from the institutional home page. Two institutions did not have library Web sites.

On each of two test dates, the 31 home pages for the libraries identified were tested twice in Watchfire's WebXACT accessibility tester at <>, once for WCAG AAA compliance and again for Section 508 compliance. Most accessibility tests require manual verification of various checkpoints. In an effort to be objective, only automated checkpoints were considered, greatly reducing the number of testable checkpoints. For Section 508, data was collected on only 3 of 16 guidelines while the WCAG was represented by 14 out of 65 checkpoints.

In addition, each page was tested against the W3C's HTML validator at <>. Sites that lacked necessary components, such as a Doctype and character encoding information, required by the validator, were revalidated in override mode with the necessary components added. The number of errors recorded reflected the Doctype and character encoding which would result in the fewest errors for the Web page. Generally, this meant using the Doctype for HTML 4.01 Transitional and the encoding for ISO-8859-1. Only one site tested had been built with XHTML 1.1 Strict and was tested against that Doctype.

Due to the lack of meaningful statistical findings in earlier studies, the data were analyzed simply by calculating the average of sites passing or failing the automated checkpoints. Since passing the automated checkpoints is not a measure of qualitative accessibility and because many bad practices can be invisible to accessibility testers, the passing of automated checkpoints merely identifies which institutions are probably aware of online accessibility testers and have made the effort to achieve some level of conformance. Each site was tested at the beginning of the fall 2003 term and again near the end of the same term to provide limited longitudinal data. The data collected and presented here is intended to represent the first year in an annual study.

Table 1: Library Web Accessibility at Kentucky's 33 4-Year Degree Granting Colleges and Universities. Data collected August 6, 2003 and December 13, 2003

Library Validity Section 508 Priority 1 Priority 2 Priority 3
  # of errors # of types of errors / total # of instances
Alice Lloyd C.  46  2 / 4  2 / 4  4 / 108  3 / 26
 46  2 / 4  2 / 4  4 / 108  3 / 26
Asbury C.  80  1 / 50  1 / 50  4 / 68  3 / 15
 80  1 / 50  1 / 50  4 / 68  3 / 16
Bellarmine U.  15  1 / 13  1 / 13  3 / 9  3 / 15
 31  1 / 25  1 / 25  3 / 10  4 / 15
Berea C.  1  PASS  PASS  2 / 33  3 / 6
 2  1 / 1  1 / 1  2 / 32  3 / 7
Brescia C.  44  1 / 43  1 / 43  2 / 5  3 / 26
 32  1 / 31  1 / 31  3 / 6  3 / 25
Campbellsville U.  47  1 / 11  1 / 8  3 / 153  4 / 21
 47  1 / 11  1 / 8  3 / 153  4 / 21
Centre C.  1  1 / 2  1 / 2  2 / 7  2 / 5
 1  1 / 2  1 / 2  2 / 7  2 / 5
Clear Creek Baptist Bible C.  118  1 / 1  PASS  5 / 14  4 / 14
 112  1 / 1  PASS  5 / 14  4 / 14
Cumberland C.  17  2 / 4  1 / 2  4 / 9  4 / 7
 14  1 / 11  1 / 11  2 / 22  3 / 10
Eastern Kentucky U.  63  1 / 1  1 / 1  2 / 46  3 / 21
 31  1 / 1  PASS 1  / 1  4 / 19
Georgetown C.  90  1 / 38  1 / 38  4 / 216  3 / 34
 108  1 / 39  1 / 39  4 / 229  3 / 36
Kentucky Christian C.  24  PASS  PASS  2 / 69  2 / 13
 29  1 / 1  PASS  3 / 78  3 / 18
Kentucky Mountain Bible C.  10  1 / 9  1 / 9  3 / 16  2 / 9
 10  1 / 9  1 / 9  3 / 16  2 / 9
Kentucky State U.  5  1 / 4  1 / 4  2 / 6  3 / 17
 22  1 / 21  1 / 21  2 / 10  3 / 17
Kentucky Wesleyan C.  20  1 / 14  1 / 14  4 / 34  4 / 12
 20  1 / 14  1 / 14  4 / 34  4 / 12
Lindsey Wilson C.  22  1 / 8  1 / 8  3 / 10  3 / 16
 22  1 / 8  1 / 8  3 / 10  3 / 16
Louisville Bible C.  n/a        
Mid-Continent Baptist Bible C.  Validator Error  1 / 3  1 / 3  2 / 5  4 / 13
 1 / 3  1 / 3  2 / 5  4 / 13
Midway C.  32  1 / 22  1 / 22  2 / 28  4 / 14
 28  1 / 19  1 / 19  2 / 15  4 / 18
Morehead State U.  262  1 / 5  1 / 4  2 / 67  3 / 16
 257  1 / 5  1 / 4  2 / 61  3 / 16
Murray State U.  57  PASS  PASS  3 / 32  4 / 16
 62  PASS  PASS  3 / 38  4 / 22
Northern Kentucky U.  VALID  PASS  PASS  PASS  PASS
Pikeville C.  64  2 / 58  1 / 56  4 / 39  4 / 21
 64  2 / 58  1 / 56  4 / 39  4 / 21
Simmons Bible C.  n/a        
Southern Baptist Theological Seminary  101  2 / 58  2 / 58  2 / 29  2 / 13
 101  2 / 58  2 / 58  2 / 23  2 / 13
Spaulding U.  70  2 / 15  2 / 15  2 / 23  4 / 15
 69  1 / 5  1 / 5  1 / 17  3 / 6
Sullivan U.  31  2 / 13  1 / 11  4 / 28  5 / 13
 64  2 / 12  1 / 10  4 / 21  5 / 9
Thomas More C.  30  1 / 3  1 / 3  4 / 36  3 / 4
 30  1 / 3  1 / 3  4 / 36  3 / 4
Transylvania U.  26  1 / 16  1 / 16  3 / 21  4 / 25
 44  1 / 5  1 / 5  4 / 64  3 / 11
Union C.  39  1 / 26  1 / 26  3 / 103  2 / 6
 39  1 / 26  1 / 26  3 / 103  2 / 6
U. of Kentucky  72  1 / 5  PASS  4 / 57  4 / 18
 78  1 / 5  PASS  4 / 48  4 / 22
U. of Louisville  29  1 / 3  PASS  3 / 60  4 / 57
 30  2 / 4  1 / 1  3 / 58  4 / 57
Western Kentucky U.  94  2 / 7  1 / 5  3 / 69  4 / 17
 93  2 / 3  1 / 1  3 / 73  3 / 13
Legend: The first row for each institution, in white, represents data collected on August 6, 2003. The second row, in light gray, represents data collected on December 13, 2003.

Looking at data from both August and December 2003, only one institution, or 3 percent of those surveyed, demonstrated the use of valid HTML and passed all automated checkpoints. Most activity was focused on Section 508 and Priority 1 issues. In August, 4 institutions out of 31, or 13 percent, passed Section 508 automated checkpoints. In December that figure was cut in half with only 2 institutions, or 6 percent, passing. The three Section 508 errors caught by automated tests were:

(a) "A text equivalent for every non-text element shall be provided,"
(i) "Frames shall be titled with text that facilitates frame identification and navigation," and
(n) "When electronic forms are designed to be completed on-line, the form shall allow people using assistive technology to access the information, field elements, and functionality required for completion and submission of the form, including all directions and cues."

Sites designed with frames, whether in violation of Section 508 guideline (i) or not, simply need to be redesigned. Guideline (a) which requires text equivalents is very easy to fix. An image is only one HTML element that can take alternative text but, since it is the most common, it makes a fine example:

<img src="image.jpg" height="10" width="20">

Above is an image without alternative text (or a text equivalent). The solution is simple:

<img src="image.jpg" alt="replacement text" height="10" width="20">

The "alt" tag value should be replacement text for the graphic, not a description of the graphic. In browsers where the "alt" tag is implemented correctly according to standards (like Mozilla 1+ and Netscape 7), the "alt" tag value is not revealed on a mouse over [51]. In the event that the graphic is a place-holder or spacer, use the "alt" tag with a null value:

<img src="image.jpg" alt="" height="1" width="1">

Keep in mind, if you are designing according to Web standards, you should not need to use spacer graphics and null "alt" tags.

The biggest problem for those who have their "alt" tags in place but still fail to meet the automated Section 508 checkpoints is the pesky <input> missing a <label>. The W3C recommends the following syntax for <label>s:

<label for="age">Your Age: <input size="10" id="age" value="Age" /></label>

If the form uses a simple layout table, it may be necessary to make a small alteration which seems to function just as well:

<label for="age">Your Age: </label><input size="10" id="age" value="Age" />

The rule is that the value of the <label>'s "for" attribute identifies the <input>'s "id" attribute. In addition, "id"s can only be used once on a Web page. This means that if you have a series of radio inputs under one question, each <input> requires a unique <label> and "id":

<p>Your Age Range: </p>
<label for="a">20-25: <input type="radio" id="a" name="age" value="20-25" /></label>
<label for="b">25-30: <input type="radio" id="b" name="age" value="25-30" /></label>
<label for="c">30-35: <input type="radio" id="c" name="age" value="30-35" /></label>

While Section 508 automated checkpoints boil down to "alt" tags and <label>s, Section 508 as a whole is quite more than that. Make sure that satisfying the automated checkpoints is just one step on the path to satisfying all Section 508 guidelines.

Priority 1 automated checkpoints saw higher compliance levels than Section 508. In August, 7 institutions, or 23 percent, passed while in December only 6 institutions, or 19 percent, passed. Automated checkpoints for Priority 1 are similar to those for Section 508 with only one exception: input labels fall under Priority 2. The only Priority 1 error that doesn't suggest a redesign is in order is missing alternative text. Taking into account that 81 percent of academic library home pages lacked alternative text in December 2003 shows that there is much room for improvement.

In December, 14 institutions of the 29 that did not pass Section 508 automated checkpoints had 5 errors or less and 4 institutions had only one error. For these institutions this could simply be the result of carelessness—good intentions without the time to follow through.


Based on low levels of compliance with the basic principles of Web accessibility and the potential of legal threats, institutions need to take accessibility issues more seriously. While the impetus to do so should fall under the rubric of professional ethics rather than avoiding a legal threat, either reason will suffice. Much like issues involving copyright, librarians embrace, on ethical terms, the protection of intellectual property while, at the same time, the threat of litigation hangs over our heads. Ethics may be important but they are also cheap. Litigation, on the other hand, is expensive. On issues such as standards, librarians understand how ignoring cataloging standards could have a negative impact on accessing collections. Can we afford to ignore Web standards when doing so impacts accessibility?

It is unfair to conclude that lack of accessibility compliance is due to blatant disregard. It is not uncommon for those who are responsible for Web development at university libraries to find their work responsibilities divided across departments and in various areas of library service. This is only a problem if these divided responsibilities do not allow sufficient time to be spent understanding and implementing Web standards and accessibility. This does require a specialist's understanding and attention if the expectation is that the work is considered professional and important.

As noted earlier in the surveys conducted by Axel Schmetzke, Web accessibility was seen to improve over time. While his surveys could have marked the growing awareness of Web accessibility at the University of Wisconsin, this growing awareness could easily have been the result of his surveys. That, simply stated, is the purpose of this current exercise. The intention has been to show where the Kentucky colleges and universities are and provide a direction for them as well as for other college and universities. If we embrace these ideas as an ethic, we could be on the way to establishing an exemplary Web presence.

Notes and References

[1] Axel Schmetzke, "Web Accessibility at University Libraries and Library Schools," Library Hi Tech 19, no. 1 (2001): 35-49. 39.

[2] Michael Burks, "Chapter 12: Emerging Technologies," Accessible Web Sites, (Glasshaus: Birmingham, UK, 2002), 324. Burks cites North Carolina State University's Center for Universal Design, Available: <> (accessed March 27, 2004).

[3] Association of Tech Act Projects, "Summary of State Information Technology Access Laws and Policies." Available: <> (accessed March 27, 2004). A thorough list of state laws, policies, and guidelines regarding accessibility that have been placed online is available at Information Technology Technical Assistance and Training Center (ITTATC) Website: <> (accessed March 27, 2004).

[4] U.S. Census Bureau, "12th Anniversary of Americans with Disabilities Act (July 26)," CB02-FF.1 (July 12, 2002). Available: <> (accessed March 27, 2004).

[5] U.S. Department of Commerce, Economics and Statistics Administration, and National Telecommunications and Information Administration, "Falling Through The Net: Toward Digital Inclusion: A Report on Americans' Access to Technology Tools," (October 2000)" Available: <> (accessed March 27, 2004).

[6] U.S. Census Bureau, "No. 18. Resident Population—States: 1980 to 2001," Statistical Abstract of the United States: 2002. 22. The figure given for the population of the US in 2001 is 284,797,000 and for Kentucky, 4,066,000. Available: <> (accessed March 27, 2004).

[7] Cathy Henderson, 2001 College Freshmen with Disabilities: A Biennial Statistical Profile (HEATH Resource Center/American Council on Education, Washington, D.C. 2001) 7. Available: <> (accessed March 27, 2004). These statistics are reported differently in Tim Spindler's article "The Accessibility of Web pages for Mid-Sized College and University Libraries," [Reference & User Services Quarterly 42, no. 2 (Winter 2002) 149-154, 149]. He writes, "In 2000, 40.4% of freshmen reported having a learning disability, 7.1% reported having an orthopedic disability, and 19.9% reported having a visual disability." These figures represent portions of those students reporting a disability, not the entire population of college and university freshmen.

[8] H. Stephan Kaye, "Computer and Internet Use Among People with Disabilities," Disability Statistics Center, Institute for Health and Aging, University of California (March 2000) 5. Available: <> (accessed March 27, 2004).

[9] Humphrey Taylor, "How the Internet is Improving the Lives of Americans With Disabilities," The Harris Poll #30 (June 7, 2000). Available: <> (accessed March 27, 2004).

[10] Valerie Lewis and Julie Klauber, "[Image] [Image] [Image] [Link] [Link] [Link]: inaccessible Web design from the perspective of a blind librarian," Library Hi Tech 20, no 2 (2002) 137-140. 138.

[11] Jeffrey Zeldman, Designing with Web Standards, (New Riders: Indianapolis, 2003): 151.

[12] Steve Noble, "Web access and the law: a public policy framework" Library Hi Tech 20, no. 4 (2002): 399-405. 401-2.

[13] Ibid., 402.

[14] Noble, 403.

[15] Ibid., 403. The text of the letter is available at <> (accessed March 27, 2004).

[16] Cynthia D. Waddell, "US Web Accessibility Law in Depth," Accessible Web Sites (Glasshaus: Birmingham, England, 2002): 329-350. 334.

[17] Ibid.

[18] Aaron P. Silberman, "Disabled Access and the Internet: Legal Issues Presented for the Bar Association of San Francisco Barristers High Tech Law Committee," (August 2002). Available: <> (accessed March 27, 2004).

[19] Ibid.

[20] Ibid.

[21] Ibid.

[22] Ibid.

[23] Architectural and Transportation Barriers Compliance Board (ATBCB), "Electronic and Information Technology Accessibility Standards." Available: <> (accessed March 27, 2004).

[24] ATBCB.

[25] Ibid. Also: 29 U.S.C. 794d. § 1194.1.

[26] Noble, 403.

[27] Ibid., 404.

[28] US Department of Education, Office for Civil Rights, Letter Docket No. 09-97-2002 (1997). Available: <> (accessed March 27, 2004).

[29] Kentucky Revised Statutes (KRS) 61.980-61.988, specifically 61.982. The definition for "covered entity" meaning "the state or any state-assisted organization" is drawn from KRS 61.980 item (4) and "access," defined in terms of Section 508, is from KRS 61.980 item (1).

[30] KRS 61.982.

[31] American Library Association, "Library Bill of Rights." Available: <> (accessed March 27, 2004).

[32] ATBCB. From the introductory material preceding the guidelines: "...the first nine provisions in §1194.22, paragraphs (a) through (i), incorporate the exact language recommended by the WAI in its comments to the proposed rule or contain language that is not substantively different than the WCAG 1.0 and was supported in its comments." Also included on the Web page is a chart to convert the Section 508 items to WCAG checkpoints.

[33] "Agenda: W3C Web Accessibility Initiative (WAI) First Technical Meeting." (May 22, 1997). Available: <> (accessed March 27, 2004).

[34] World Wide Web Consortium (W3C), Web Content Accessibility Guidelines 1.0 (WCAG). Available: <> (accessed March 27, 2004).

[35] Zeldman, 21. WaSP is available online at <> (accessed March 27, 2004).

[36] Ibid., 420-2.

[37] Ibid., 21.

[38] The W3C provides a validator for HTML at <> and one for CSS at <> (accessed March 27, 2004).

[39] Jukka Korpela's "Improving accessibility with accesskey in HTML forms and links." Available: <> (accessed March 27, 2004).

[40] Ibid. There are de facto standards for implementing numerical accesskeys. A good place to start defining accesskeys is Mark Pilgrim's "Accessibility Statement," Dive into Accessibility. Available: <> (accessed March 27, 2004).

[41] Erica B. Lilly and Connie Van Fleet, "Wired But Not Connected: Accessibility of Academic Library Home Pages," The Reference Librarian 34, nos 67/68 (1999): 5-28.

[42] Tim Spindler, "The Accessibility of Web Pages for Mid-Sized College and University Libraries," Reference & User Services Quarterly 42, no. 2 (2002): 149-154.

[43] Schmetzke (2001).

[44] Ibid., 44.

[45] Ibid.

[46] Axel Schmetzke, "Web Accessibility Survey Site," Available: <> (accessed March 27, 2004).

[47] Sue Polanka and Jack O'Gorman, "Guidelines for Creating Accessible Library Web Pages," Internet Reference Services Quarterly 5, no. 3 (2001): 51-57; Sean Kinder, "Web Page Accessibility for Users with Disabilities," Kentucky Libraries 65, no. 4 (Fall 2001): 21-22; and Cheryl H. Kirkpatrick, "Getting Two for the Price of One: Accessibility and Usability," Computers in Libraries 23, no. 1 (January 2003): 26-29.

[48] Mark Pilgrim, "Dive into Accessibility." Available: <> (accessed March 27, 2004).

[49] Joe Clark, Building Accessible Websites (New Riders: 2003).

[50] Zeldman.

[51] Ibid., 342.

Copyright © 2004 Michael Providenti

Top | Contents
Search | Author Index | Title Index | Back Issues
Previous Article | Next article
Home | E-mail the Editor


D-Lib Magazine Access Terms and Conditions