The Accessibility of Online Library Resources for People with Print Disabilities: Research and Strategies for Change

Axel Schmetzke

University of Wisconsin-Stevens Point, Library, 900 Reserve Street,

Stevens Point, WI 54481, USA

aschmetz@uwsp.edu

Abstract. As more and more learning in higher education takes place in cyberspace, the accessibility of online resources to all people, including those with disabilities, has become an important issue. This paper reviews the research on the accessibility of library web sites and selected online information resources, and it discusses the role that future research should play in facilitating an inclusive learning environment.

1 Introduction

Recent developments in telecommunication—particularly the coming of age of the internet—have had a strong impact on higher education. Over the past decade, the way information is disseminated in the campus environment has undergone drastic changes. Increasingly, print-based information is being substituted with its digital equivalent. Today, the web, along with email, provides the main, if not the sole, channel for a variety of education-supporting resources: official campus web pages with crucial administrative information, class syllabi, course readings, and web-mediated distance education programs. Libraries are particularly strongly affected by this digital revolution. The shift from the physical to the virtual permeates almost every aspect of their operation. There is hardly a single type of library resource that has not shifted, at least to some extent, to a digitized, web-based format. Catalogs, indexes and full-text article databases, dictionaries and encyclopedias, e-books and e-journals, reserve materials, virtual reference services as well as information about the library itself are now commonly accessed through library web sites.

With the growing importance of digitized, web-based information, the issue of access to information is no longer limited to the physical realm. Just as there are enabling and disabling conditions in the physical environment, so are there conditions in cyberspace that result in the inclusion or exclusion of people. To some extent, the ability to access web-based information is a question of the proper assistive technology, such as a modified computer keyboard, an enlarged screen display, or a properly configured screen-reading program. However, assistive technology alone cannot overcome the barriers that are created at a more basic level: the format in which content is presented. If not properly formatted, or designed, online resources are not accessible to people with certain disabilities—no matter how advanced and plentiful the assistive technology available to them may be.

While the notion of universal design has been discussed extensively in the architectural and exterior/interior design literature, its application to the electronic environment was, until a few years ago, rarely addressed in traditional print media. Instead, the theme was mainly carried by a rather tightly knit network of dedicated people who gathered at disability- and web-related conferences and shared their insights in form of presentations, white papers and web-posted articles. It was not until 1996 that accessible web design emerged as an issue in the professional library literature. The past three years have seen a noticeable increase in library-related journal publications that seek to raise awareness concerning the need for accessible web design and provide practical tips [1, 2]. In 1998, researchers began collecting data on the accessibility of library web sites. The first studies on the accessibility of online library resources (other than web pages) will appear shortly in two special-theme issues of Library Hi Tech (2002, Vol. 20, Issues 2 and 4), guest-edited by this author. Methodology and findings of these studies will be described in the following sections.

2 Web Site Accessibility Studies

24 studies investigating the accessibility of web sites are known to this author. Of these, 10 include library web sites. Most of the institutions targeted are in the United States. Caution is advised in extrapolating from these findings to the situation in other countries.

2.1 Methodology

With very few exceptions, all studies relied exclusively on the automated checking capabilities of “Bobby,” a tool created by the Center for Applied Special Technology (CAST) to help people evaluate the accessibility of their web pages [3]. The downloadable version of Bobby, which runs as an application on a personal computer, is capable of testing larger sets of web pages on a given web site. For each page checked, Bobby provides information pertaining to the type, number, and location of accessibility errors—both minor and major ones. Bobby also issues a summary report for each set of web pages.

Within the context of this article, the term “Bobby-approved” is used, in a rather lax manner, to indicate that no major access barriers (“Priority 1” errors—in the terminology of the 1999 Web Content Accessibility Guidelines established by the World Wide Web Consortium’s Web Accessibility Initiative) were detected by Bobby’s automated function. Strictly speaking, this use of the term is incorrect: Full Bobby approval would also require a “manual” evaluation of those Priority 1 items that Bobby cannot automatically check.

The exclusive use of Bobby’s automated function for evaluative purposes is also problematic for other reasons, which this author discusses elsewhere in much detail [1, 2]. Suffice it to say here that Bobby testing will result some falsely negative and positive findings. Despite such shortcomings, Bobby is a good evaluation tool in studies where the accessibility of hundreds or thousands of individual web pages are evaluated and a rough measure of accessibility will do.

2.2 Findings

Table 1 summarizes the data yielded by library-related web studies. Average accessibility in the various library data sets ranges between 19% and 75%, and the average number of errors per page varies between 1.3 and 6.1. Web accessibility tends to be higher at academic libraries than at public libraries. While web accessibility has improved in some pockets (University of Wisconsin libraries), larger data sets show no progress. Perhaps most remarkable is the fact that the web pages of North American library schools tend to be particularly inaccessible. It is reasonable to assume that such disregard for accessible web design reflects not only the attitude of web designers but also that of the library school faculty and staff, who hire the designers and give them direction. The low web accessibility at library schools also has another consequence. Students with print disabilities have a lower chance of successfully passing through library school programs.

Table 1. Average accessibility of library-related web sites, by type of institution and study, in terms of the percentage of accessible web pages and errors per page

Study/Data Set	Year	Geographic Focus	Accessibility- Home Page	Accessibility- Top Layer
			%	%	errors/page
Academic Libraries
Blake [4]	2000	Arkansas, U.S.	37	–	–
Craven [5]	2000	U.K.	37	–	–
Lilly & Van Fleet [6]	1998	U.S.	40	–	–
Schmetzke [7]	2002	Wisconsin, U.S.	64	75	1.3
Schmetzke [7]	2001	Wisconsin, U.S.	69	43	6.1
Schmetzke [7]	2000	Wisconsin, U.S.	54	40	2.9
Schmetzke [7]	1999	Wisconsin, U.S.	–	31	3.6
Schmetzke [1]	2002	U.S. (49 sites)	51	47	5.0
Schmetzke [1]	2002	Canada (7 sites)	43	53	5.9
Schmetzke [1]	2002	U.S. (24 sites)	71	53	4.2
Schmetzke [2]	2000	U.S. (24 sites)	71	59	2.8
Yu [8]	2001	California, U.S.	38	–	–
Public Libraries
Kester [9]	1999	N. Carolina, U.S.	21	–	–
Lilly & Van Fleet [10]	2000	U.S.	19	–	–
Ormes & Peacock [11]	1999	U.K.	<31	–	–
Library Schools
Schmetzke [1]	2002	U.S (49 sites)	31	33	8.0
Schmetzke [1]	2002	Canada (7 sites)	14	36	5.0
Schmetzke [1]	2002	U.S. (24 sites)	38	30	10.1
Schmetzke [2]	2000	U.S. (24 sites)	38	23	8.0

3. Accessibility of Online Resources Other than Web Pages

To the best of the author’s knowledge, the only studies on the accessibility of online library resources other than web pages are those soon to appear in two special-theme issues of Library Hi Tech (Vol. 20, Issues 2 and 4, 2002). This section contains a selective overview of the findings, pertaining to the accessibility of online catalogs, aggregated journal indexes/databases, e-journals and reference works. Space limitations do not permit to include the findings of the second special-theme issue, which will include discussions of electronic reserve systems and courseware.

3.1 Methodology

The methods to evaluate accessibility varied from study to study. Two major approaches were used: compliance checking and functionality testing employing assistive technology. For the former, authors selected major search-related screens and then checked their design for compliance with established accessible design principles, such as the W3C/WAI guidelines or the Access Board standards issued in connection with Section 508 of the U.S. Rehabilitation Act Amendment of 1998. This compliance check was done either “manually” or with the aid of an automated evaluation tool, such as Bobby. For functionality testing, authors selected some basic search-related tasks and then attempted to perform these tasks using assistive technology (typically screen readers, such as Jaws and Window-Eyes). Some authors, of whom at least one was blind, did these searches themselves; others collaborated with sight-impaired users already familiar with the assistive device. With one exception, all studies focused exclusively on search-related pages or the pages that function as a gateway to full-text material. Only one author, Coonin, who assessed the accessibility of e-journals, also looked at the format of the full-text material.

It must be pointed out that the findings reported below should be treated with caution. The interfaces of some finding tools, such as the Voyager online catalog, can be customized to some extent by library staff. Accessibility of the overall system is therefore also affected by the customized features selected by individual institutions. Furthermore, the data reported here, which were collected in spring and summer of 2001, may no longer be current.

3.2 Findings

Online catalog: Two online catalogs were studied. Axtell and Dixon looked at the Endeavor product Voyager, an integrated library management system widely used in North American Academic Libraries [12]. Focusing on Voyager’s public access component, the researchers arrived at the overall conclusion that, with some effort, the system “can be successfully navigated by experienced users of assistive technology.” They recommend, among others, the following improvements:

· Add a mechanism on each page to “skip navigation links.”

· Convey the association among a group of buttons not by visual cues alone.

· Modify Javascipt code in the “sort by” list box.

· Add label tags to the range buttons on the limit screen.

Johns looked at Epixtech’s iPAC 2.0, an online catalog system more likely to be found at public libraries [13]. She concludes that “many accessibility features are currently coded and provide solid accessibility in the product. Johns recommends, among others, to improve the logical order for tables, lengthy displays, sequences of links and split screens, and to be more consistent when using elements such as buttons, edit boxes and combo boxes. She was pleased that the product developers were open to constructive criticism and eager to improve their product.

Indexes and databases: Five studies examined the accessibility of online indexes and databases, including those with broad-range coverage, such as Ebscohost and Proquest Research Library, and those that are subject-specific, such as Medline and Cancerlit. The findings are depicted in Table 2. The categorization of the evaluation results in this table—a five point scale ranging from “very accessible” to “absolutely inaccessible”—were established by this author. Most authors of the original studies agreed with this categorization scheme and the placement of their findings within it (i.e., this author’s interpretation of their research findings in terms of his descriptive categories).

Table 2. Accessibility of indexes/databases

Index/Database	Evaluation Results*
	Very accessible	Mostly minor problems; mildly reduced accessibility	Accessibility significantly reduced	Major problems; severely reduced accessibility	Absolutely inaccessible
Ebscohost		Bowman [14] Horwath [15]	Riley [16]
Proquest Research Lib.			Bowman [14]
Lexis-Nexis Acad. U.		Bowman [14]
OCLC FirstSearch		Byerley et al. [17]	Riley [16]
InfoTrac Web (Gale)		Byerley et al. [17]	Riley [16]
Ovid	McCord et al.  [18]
Cancerlit		McCord et al. [18]
Electric Library Plus			Horwath [15]
Hazardous Sub.  Data B.		McCord et al. [18]
MSDS--Cornell U.		McCord et al. [18]
MSDS--U. of Vermont		McCord et al. [18]
MedlinePlus (NIH)		McCord et al. [18]
PubMed			McCord et al. [18]
Toxline		McCord et al. [18]

As Table 2 reveals, ten of the fourteen resources studied are included in the “mildly reduced accessibility” category. For the three (out of these ten) resources that were evaluated by more than one research team, the findings are typically close but not identical. Riley found Ebscohost, OCLC FirstSearch and Infotrac (Gale) to be less accessible than other research teams. Altogether, six of the fourteen resources are included in the ”significantly reduced accessibility” category. While only one index/database interface is described as “very accessible,” none of the resources fall into the “severely reduced accessibility” or “absolutely inaccessible” categories.

Reference Resources: Horwath’s study includes the electronic versions of two traditional reference resources: the Encyclopedia Britannica Online and the Oxford English Dictionary (OED) Online [15]. She found only the former to be very accessible. The OED contained barriers that significantly reduced its accessibility. The accessibility of the Encyclopedia Britannica Online comes as a surprise to this author. When asked about its accessibility for people with disabilities at an exhibition during the American Library Association Conference in 2000 (summer), the publisher’s sales representative denied, rather briskly, any knowledge about this aspect of this product and discouraged any discussion about the topic.

E-Journals: Coonin [19], who examined the accessibility of e-journals of eleven providers (see Table 3), found, with one exception (Kluwer Online), some major accessibility errors in the “basic access pages” (introductory, basic search or result pages) of all e-journals. Lack of alternative tags for images was the most frequent problem. Alternative tags for image-map hot spots were missing among the pages of four of the e-journals. Four e-journal providers neglected to title all the frames.

Table 3. Accessibility of e-journals by provider: number of major accessibility errors in “basic access” pages and electronic file format of e-journal articles (Coonin [19])

E-journal provider	Major Accessibility Errors	Electronic Format of E-Journal Articles
		HTML	Text PDF	Image PDF	Other
BioOne (SPARC)	5	X	X
Catchword	3		X		RealPage
Emerald	3	X	X
HighWire*	4	X	X
IDEAL	3		X
JSTOR	4			X
Kluwer Online	0		X
Project MUSE**	4	X	X	X
Science Direct	4	X	X
SpringerLink	4	X	X
Wiley Interscience	5	X	X

*Some journals in text PDF only. **Some journals in image PDF only.

As indicated earlier, Coonin also took notice of the format in which the articles themselves are provided. Six of the providers offer an accessible html format option for their e-journal. Of the five providers who do not offer this option consistently, three utilize a text-based PDF format, which, for the most part, is an accessible product. All of JSTOR’s, and some of Project Muse’s e-journal articles, are provided in image-only PDF files, which cannot be accessed with screen readers. Systems, such as Kurzweil 1000, are emerging on the market that are capable of converting the image-based text information into ASCII text with the aid of optical character recognition (ORC) software, but the converted product contains errors and is thus considered by many to be an unacceptable long-term solution.

4. Discussion: Research as an Instrument of Change

The research findings reviewed in the previous two sections reveal a mixed picture: Even at academic libraries, where web site accessibility tends to be higher than at public libraries, people using assistive technology often encounter access barriers. On the positive side, few of the online resources reviewed—catalogs, indexes/databases, reference works, and e-journals—were absolutely inaccessible. For most of these, accessibility was reduced—but not to the point of absolute inaccessibility. While the above picture is not as grim as it could be, much needs to improve to make library resources fully accessible to all users. It is particularly disturbing that library schools do not appear to take on the leadership role that they should. If the low accessibility of library schools’ web pages is any indication of the level of awareness at these institutions, new librarians will not enter the profession with the knowledge and attitudes conducive to the creation of a barrier-free online environment.

Accessibility research may help to bring about change in three ways: First, by providing a general picture of where libraries are and where they need to go, it gives impetus to the rewriting of existing policies in more inclusive terms and to putting into place adequate mechanisms for their implementation. Second, to the extent that web site accessibility studies identify individual institutions, those with low accessibility may become motivated to improve their “ranking.” No institution likes to stick out negatively in comparative studies. Also, with the publication of these data, institutions that do not respond to documented access deficiencies may find themselves in a particularly vulnerable position should legal complaints be filed against them (as it may happen, for example, in the U.S. under the Americans with Disabilities Act). Third, information about the accessibility of commercial products is badly needed by librarians making procurement decisions. Most librarians have no idea whether the many electronic resources to which their libraries subscribe, or the resources that they are considering for purchase, are accessible to people using assistive technology. The reason for this is simple: There is hardly any information out there. Product vendors are only beginning to understand the issue, and those who do provide information, tend to couch it in vague phrases (“compliance with the Americans with Disabilities Act”) and are typically unable to answer more specific questions, such as questions about the methods used to evaluate the accessibility of their products. While, thanks to Bobby and other such tools, it is relatively easy to get a rough idea about the accessibility of ordinary web pages, the evaluation of dynamic interfaces exceeds the expertise (and time constraints) of most librarians. The studies soon to be published in Library Hi Tech constitute a first step towards filling this information vacuum. Hopefully, these issues will stimulate further evaluative research in this area, help librarians to make informed procurement decisions, and provide information that may encourage vendors to develop truly inclusive products.

Laudable as the efforts behind these special theme-issues undoubtedly are—to some extent they are likely to serve the intended purpose—they constitute just a first step. Ultimately, a more elaborate system to monitor and encourage accessibility of online library resources needs to be put in place—a system that generates a continuous flow of up-to-date information. Electronic resources are moving targets. New versions come out faster than evaluative research can be disseminated through traditional venues (journals and books). A few librarians, when faced with yet another product to be considered for purchase, may get inspired by the published research and copy its methodology to conduct their own accessibility evaluations. However, most librarians will not have the time or expertise to do so. What is needed is an information source similar to that of Consumer Reports. A project to create such an information source is currently under consideration by this author: Tentatively dubbed “Project ADOPT-IT”, it would take the following shape: Small groups of librarians are formed of which each adopts a particular product category (electronic reserve, virtual reference, e-books, etc.). Each group evaluates the accessibility of products in the adopted category as they emerge on the market and shares its findings through a web-based clearinghouse created for this particular purpose. Any library considering purchasing a particular online information resource would be welcome to check the ADOPT-IT clearinghouse for accessibility-related information.

Project ADOPT-IT may not work out for one or the other reason. However, this should not keep supporters of inclusive information technology from exploring other, more viable approaches. As this study suggests, progress is slow at best. Over half of the library web sites are still fraught with access problems, and judged by their own web pages, library schools are slow to recognize the problem. Without continuous efforts to educate, to reshape institutional policy, and to challenge vendors to make more accessible products, the web is unlikely to take on the quality that Burners-Lee, its inventor and the current director of the World Wide Web Consortium, had in mind when he created it: “The power of the Web is its universality. Access by everyone regardless of disability is an essential aspect” [20].

References

1. Schmetzke, A.: Web Accessibility at University Libraries and Library Schools: 2002 Follow-Up Study. In Hricko, M., (ed.): Design and Implementation of Web-Enabled Teaching Tools. Idea Group Publishing, Hershey, Pennsylvania (in press).

2. Schmetzke, A.: Web Accessibility at University Libraries and Library Schools. Library Hi Tech. 19:1 (2001) 35-49.

3. Bobby. CAST. 22. Accessed April 2002. http://www.cast.org/Bobby/.

4. Blake, S. Universal Access, the ADA, and Your Library Web Page. Arkansas Libraries. 57:1 (2000) 19-24.

5. Craven, J.: Electronic Access for All: Awareness in Creating Accessible Web Sites for the University Library. Disability and Information Systems in Higher Education (DISinHE). 18. Oct. 2000; accessed 21. April 2002. <http://www.disinhe.ac.uk/library/article.asp?id=34>.

6. Lilly, E.B., van Fleet, C.: Wired But Not Connected: Accessibility of Academic Library Home Pages. The Reference Librarian. No. 67/68 (1999) 5-28.

7. Schmetzke, A.: Web Accessibility Survey Homepage. 10. April 2002. http://library.uwsp.edu/aschmetz/Accessible/websurveys.htm.

8. Yu, H.: Web Accessibility and the Law: Recommendations for Implementation. Library Hi Tech. 20:4 (2002) in press.

9. Kester, D.: Measuring the Sight of Your Web Site. North Carolina Libraries. 57:3 (1999) 114-117.

10. Lilly, E.B., van Fleet, C.: Measuring the Accessibility of Public Library Home Pages. Reference & User Services Quarterly. 40:2 (2000) 156-163.

11. Ormes, S., Peacock, I.: Virtually Inaccessible? Making the Library Virtually Accessible. LTWorld. 3. Feb, 1999; accessed 21. April 2002. <http://www.sbu.ac.uk/litc/lt/1999/news1317.html>.

12. Axtell, R., Dixon, J.M.: Voyager 2000: A Review of Accessibility for Persons with Visual Disabilities. Library Hi Tech. 20:2 (2002) in press.

13. Johns, S.M.: Viewing the Sunrise: iPAC 2.0 Accessibility. Library Hi Tech. 20:2 (2002) in press.

14. Bowman, V.: Reading between the Line: An Evaluation of WindowEyes Screen Reader as a Reference Tool for Teaching and Learning. Library Hi Tech. 20:2 (2002) in press.

15. Horwath, J.: Evaluating the Opportunities for Expanded Information Access: A Study of the Accessibility of Four Online Databases to People who are Blind and Visually Impaired. Library Hi Tech. 20:2 (2002) in press.

16. Riley, C.A.: Libraries, Aggregator Databases, Screen Readers, and Clients with Disabilities. Library Hi Tech. 20:2 (2002) in press.

17. Byerley, S.L., Chambers, M.B.: Accessibility and Usability of Web-Based Library databases for Non-Visual Users. Library Hi Tech. 20:2 (2002) in press.

18. McCord, S., Frederiksen, L., Campbell, N.: An Accessibility Assessment of Selected Web-Based Health Information Resources. Library Hi Tech. 20:2 (2002) in press.

19. Coonin, B.: Establishing Accessibility for E-Journals: A Suggested Approach. Library Hi Tech. 20:2 (2002) in press.

20. Burners-Lee, T.: Quote retrieved April 25, 2002, from the Web Accessibility Initiative (WAI) web site at http:www.w3.org/WAI/.