MDA 2004 Conference:A Holistic Approach To Web Accessibility
MDA 2004 Conference:The World Wide Web provides access to information in a manner which could not be envisaged by many little more than a decade ago. The Web also provides access to resources by people who may traditionally have how it difficult to access information such as people with visual impairments and other disabilities. The Web Accessibility Initiative (WAI) has been established to raise awareness of universal access and to develop guidelines which can help to ensure that Web pages are widely accessible.
Despite WAI's high profile and laudable aims many Web resources appear not to be universally accessible. However rather than regarding this as a failure on the part of creators of Web resources in this paper the author argues that the failures may, in part, reflect the difficulties of complying with the guidelines and the limitations in Web software applications but also may be due to limitations in the guidelines themselves.
Despite these limitations the author strongly supports the aims of universal accessibility. However rather than relying solely on the WAI guidelines the author recommends the adoption of a holistic approach to accessibility which is informed by the WAI guidelines but uses them in a wider context which reflects real world issues such as resources, technological and cultural issues and seeks to focus on the needs of the user.
The importance of universal access to Web resources is widely acknowledged. The W3C WAI (Web Accessibility Initiative) has been tremendously successful in raising awareness of the potential of the Web for providing access to resources to people with disabilities, and in producing a series of guidelines which can help to ensure that Web resources are widely accessible. The guidelines include the Web Content Authoring Guidelines WCAG [1], the Authoring Tool Accessibility Guidelines ATAG [2] and the User Agent Accessibility Guidelines [3].
The WCAG guidelines of most relevance to creators of Web sites. These guidelines provide advice on how Web pages should be developed in order to ensure that the content is accessible to people with disabilities. The guidelines state, for example, that images must be accompanied by alternative text. This will ensure that if images aren't displayed or the users cannot see images, that browsers (such as speaking browsers) will be able to process information such as a textual description of an image.
The two other guidelines (Authoring Tool Accessibility Guidelines and the User Agent Accessibility Guidelines) are aimed at the software development community. For example the User Agent Accessibility Guidelines recommend that user agent software (such as Web browsers) should be capable of processing alternative text supplied with images and the Authoring Tool Accessibility Guidelines recommend that HTML authoring tools should be capable of creating the alternative text.
From this brief summary of the key WAI documents it should be apparent that the WCAG guidelines are of direct relevance to creators of Web resources.
W3C WAI has been successful in raising awareness of the importance of universal Web accessibility and the relevance of the WCAG guidelines for use by Web authors. Many organisations include compliance with the WCAG guidelines in their organisational guidelines and policies. In addition compliance with the WCAG guidelines is increasingly being mandated by funding bodies within the public sector. For example within the UK government guidelines require compliance with WCAG AA guidelines [4] and the NOF-digitise programme also requires compliance with WCAG A guidelines [5].
Despite the high-profile of the WAI WCAG guidelines, in practice many organisations appear to experience difficulties in achieving compliance. The author has carried out several surveys of Web sites in particular communities in the UK, including the higher and further education sectors and has recently extended this work to the museums, libraries and archives sector. The surveys have made use of the Bobby accessibility checking tool [6]. This tool provides a simple way of performing partial accessibility tests on public Web sites. It must be noted, however, that use of automated accessibility testing tools such as Bobby cannot prove that a page is accessible - manual accessibility testing is required in order to do this. The best an automated tool can do is to prove that a page is inaccessible or may be accessible.
Despite such limitations, Bobby has been used to measure the extent of pages which are inaccessible and which may be accessible in a number of sectors. A survey of the accessibility of UK University home pages was published in September 2002 [7], when the UK DDA (Disability Discrimination Act) was extended to include the educational sector. The survey found that of 162 institutions only four appeared to comply with WCAG AA guidelines (and subsequent analysis reduced this by one). However 70 institutions appeared to comply with WCAG A guidelines.
This survey was repeated in July 2004 [8]. The number of institutions which appeared to comply with WCAG AA had risen from 4 to 7, and the number which appeared to comply with WCAG A had risen from 70 to 93.
From these findings it would appear that compliance with WCAG A guidelines is achieved by about half of the institutions, but compliance with WCAG AA guidelines by only a handful. These findings appear dispiriting, especially as the survey addresses only the institutions' home page and does not analyse entire Web sites.
These findings appear to be replicated in a more comprehensive survey commissioned by the Disability Rights Commission (DRC). The survey was based on the analysis of 1,000 Web sites. The findings, which were published in the Formal Investigation Report: Web Accessibility [9] indicated that 81% of the Web sites failed to comply with WAI WCAG level A guidelines.
The publication of the findings of the initial survey of the UK Higher Education sector and the DRC report led to discussions on mailing lists used within the sector and at subsequent events attended by the University Web management community. Although the work of WAI is very much appreciated within the community and there is a willingness to implement standards and best practices which will help ensure universal accessibility a number of concerns were raised with the guidelines:
Such concerns have led, however, to a feeling of uncertainty. If implementation of the WCAG guidelines is problematic, what should the Web developer who wishes to provide an accessible Web site do?
UKOLN [10] in conjunction with the JISC-funded TechDis advisory service [11] have been working together for several years in providing several workshops for the higher and further education communities. They have encountered many of the issues raised above during their work. These experiences led to the development of the TechDis/UKOLN holistic approach to e-learning accessibility, which has recently been accepted for publication [12]. This paper also had valuable input from a software developer in the e-learning unit at the University of Bath which was based on implementation experiences gained in development of accessible e-learning services.
Our work was based on the need to provide advice on accessibility for e-learning resources. We were finding that although institutions could understand the approaches needed to ensure that information resources were accessible (even if implementing such approaches was not necessarily easy) it was difficult to develop strategies for ensuring that e-learning environments were accessible. For example, providing additional information such as alternative text for images on informational resources is beneficial to all; however when developing e-learning resources in which students are expected to give answers to questions based on their understanding and interpretation of images, to provide equivalent information using alternative text has the danger that the answers to the questions may be revealed.
The e-learning community is also active in exploiting new technologies in new and challenging ways - technologies such as streaming media, 3D visualisation and immersive environments are currently being deployed in e-learning contexts. Such environments are making use of leading-edge technologies in which there may be very limited experiences in achieving accessibility.
The approach we have developed seeks to focus on the needs of the learner rather than necessarily seeking to achieve compliance with a fixed set of guidelines. The approach recognises that there may well be resource implications in achieving compliance with accessibility guidelines and other factors such as limited expertise, etc. Most importantly we feel that we should be seeking to achieve an accessible learning experience and accessible learning outcomes and not necessarily accessible e-learning resources.
Our model is illustrated in Figure 1.
Figure 1: The TechDis/UKOLN Holistic Model For E-Learning Accessibility
It should be noted that this model is built on top of the WAI work. The WAI WCAG guidelines will still need to be addressed and considered, but alongside other criteria. It should also be noted that the model requires a quality assurance framework to be deployed which requires the provision of appropriate policies and systematic procedures for ensuring compliance with the policies.
We feel that this model could be extended to address the provision of digital services within the cultural heritage sector. Many services in the sector will, of course, have an e-learning element. In addition we argue that organisations may wish to consider ensuring that digital cultural heritage services for which accessibility is difficult or costly to provide should consider provide equivalent experiences. A could example of this was illustrated following a presentation at UKOLN's Public Library conference in March 2004. Following a demonstration of a game developed using Flash which was aimed at young children, in response to a question concerning the accessibility of the game, the response was that although the game appeared popular, the library would probably have to take it down as it does not comply with WAI guidelines. However using our holistic approach we would argue that there is a need to provide an alternative experience. Since the aim of the game was to keep young children entertained while their parents use the library, an equivalent experience could be a real-life game, such as a bouncy castle or building blocks.
Another example which illustrates an approach to the development of a cultural heritage service which is focussed on the needs of a particular user community rather than aiming to follow the underlying WAI philosophy of 'universal design' is the i-Map service at Tate Modern [13]. i-Map provides an e-learning experience for paintings by Picasso and Matisse. I-map has been developed specifically for the needs of the visually impaired, including the congenitally blind. This service is based on inclusive design rather than universal design as its underlying principle for this service.
On reflection we should not be too surprised that the attempt by WAI to provide a simple set of guidelines which could help to achieve universal accessibility has its flaws. This is partly due to the immaturity of the guidelines - these are still version 1.0 and experienced IT users often feel it is sensible to wait for version 3 of a standard or application, by which time implementation experiences should have helped remove many of the problems. However a more fundamental issue is whether a universal set of guidelines is applicable for addressing the human-machine interface. The TechDis/UKOLN Holistic Model feels that the WAI guidelines have relevance in some areas but in the wider context there will be need to address issues which are not reducible to a universal checklist approach.
In this paper we have described the valuable role played by WAI in providing guidelines which can help Web developers to create accessible Web resources. The poor compliance with some of the WAI guidelines may be due to the nature of the guidelines rather than a lack of inclination to implement them. Although the aims of WAI are to be applauded, we argue that there is a need to take a broader view of accessibility than is provided in the WAI WCAG guidelines. We feel that the holistic approach outlined in this paper will address some of these concerns and should be applicable for use in a cultural heritage context.