This article is from the May 2003 issue of Update.
There are moral, economic and legal reasons for ensuring a
website does not exclude any users, whether through a disability, a particular
need or hardware/software reasons, as an article on web accessibility and design
in the January edition of Update1
pointed out.
In the article, Juliet Owen provided some simple steps to help
people ensure their websites can be accessed by as many people as possible. But
if the accessibility message is to reach everyone (and not just preached to the
converted), awareness-raising activities (articles, conference papers, seminars)
are vital.
The question of awareness levels was explored in a supporting
study for DISinHE (now TechDIS), undertaken by the Centre for Research in
Library and Information Management (Cerlim) between 1999 and 2000.2
An email survey was sent to 100 UK higher education library website developers
(taken from the contact details on the library home page) with a set of
questions to help identify whether university library web page designers were
aware of accessibility issues and, if so, whether best practice was being
followed. Questions relating to accessibility checking methods were asked, as
well as how useful the designers found them. Finally, bearing in mind that
contact details were generally provided via the library home page, respondents
were asked about the level of user input which was taken into consideration when
developing the site.
There was only a 30 per cent response rate to the survey. This
was adequate to provide indicative results, but was still disappointing. It was
also tempting to make assumptions about the reasons for the poor response, as
each site was tested for accessibility (using automated tools), and this
revealed many accessibility errors, such as poorly contrasting colours and a
lack of alternative text for graphics. Among the 30 per cent who did respond,
there was evidence of general awareness of accessible web design issues.
However, it was clear there needed to be more development in this area,
including general awareness-raising events and training in the design of
accessible websites, not just for the staff involved, but for managers and
policy decision-makers.
Thankfully, awareness does appear to be growing, due in part to
an increase in articles and papers on accessibility issues, spurred on by
legislation such as the Disability Discrimination Act 1995 and the Special
Educational Needs & Disability Act 2001. It is hoped that interest as well
as awareness is growing, as information providers and web designers realise that
it just makes good sense to try and ensure all users can enjoy the benefits
offered by web-based information resources.
A copy of the DCMS/Share the Vision publication Library
Service Provision for Blind and Visually-impaired People: a manual of best
practice3
was sent to all UK public and academic libraries and all academic departments
offering library and information studies courses. It was also available to
download from the National Library for the Blind website.4
The manual focuses on service provision for people with visual impairments, but
sections on accessible web design (Chapter 14) and on assistive technologies
(Chapter 13) should be of interest to anyone involved in the design, development
or selection of accessible websites.
Another example of help in this area is the RNIB’s Campaign for
Good Web Design.5
This works on many levels to promote universal design. It offers advice and
accessibility audits to banks, shops and public services, as well as working
with commercial web design companies such as Adobe and Macromedia to ensure that
mainstream web authoring tools are offering accessibility options.
Accessible web design does not have to be boring design. As
Owen points out, ‘accessibility does not have to compromise the visual design of
a site’. One of the key messages of the RNIB campaign is that ‘accessibility
should not stifle innovation’. This was brought to life at the recent Visionary
Design Awards, run by the National Library for the Blind. Web publishers and
designers were invited to submit their websites, which were judged on their
level of accessibility for all forms of access technology, taking into account
the experience a visually-impaired user might have if visiting the site for the
first time. Winners included: Whichbook net;6
Local Heritage Initiative;7
and Guardian Unlimited.8
Further details about this award can be found on the NLB website.9
Test for usability as well as accessibility
If guidelines and recommendations for accessible design are
followed (or alternatives offered when absolutely necessary), interesting,
interactive, accessible sites can easily be created, or inaccessibility errors
corrected, often at little cost. However, there are a number of other
considerations that must be taken into account before a site truly meets the
requirements of universal access. Owen touches on this area: ‘Following
guidelines is not enough to produce a site that is fully accessible for the
intended audience. It is useful to talk to your users.’ Even if you follow
accessibility recommendations, the importance of involving real people cannot be
stressed enough.
Involving users in accessibility testing is often referred to
as testing the usability of a site. How actual users interact with the site is
looked at rather than using an automated accessibility checking tool. A simple
explanation of the difference between accessibility and usability is described
by Howell10
as:
- accessible = technical guidelines (e.g. Web Accessibility Initiative (WAI)
Web Content Accessibility Guidelines);11
- usable = experiential (user testing to complete a task).
One complements the other and both should, wherever possible,
be conducted on a regular basis.
Usability issues were the focus of the Non-Visual Access to the
Digital Library (NoVA) project,12
funded by Resource and undertaken by Cerlim between June 2000 and July 2002.
Following Cerlim’s previous work on the Resources for Visually-Impaired Users of
the Electronic Library (Reviel) project, which included feedback from a group of
visually-impaired people on the accessibility of a variety of online resources,
it became apparent that navigation is a major problem within digital library
resources. The NoVA project was subsequently undertaken to explore the usability
of web-based resources, with a particular focus on blind and visually-impaired
users. The aim was to develop an understanding of the searching or browsing
behaviour of users who could not read or interact with a screen without the aid
of assistive technologies (such as a screen reader or screen magnification) or
having to be very near the screen.
Usability experiments
A sample of 20 sighted and 20 blind and visually-impaired
people helped with a series of usability experiments with searching and
retrieval tasks, and subsequent use of digital content. Four information-seeking
tasks were set using four web-based resources: a search engine; a library Opac;
an online shopping site; and a directory of internet resources. Tasks were
consistently set so that comparative analysis could take place between the
sighted and visually-impaired users.
All the resources chosen were, to a greater or lesser degree,
accessible according to the WAI guidelines. This was done because the purpose of
the usability testing was not to highlight a lack of accessibility features, but
rather to test sites that should — in terms of accessibility checkers — be
reasonably accessible.
All the resources chosen did, however, display degrees of
parallelism in their design. This refers to the use of tables and frames to
enable the user to perform complex selections across categories (search
facilities, navigational links, blocks of text, graphics, etc) all in one page.
Usually such designs assume visual capabilities, as when a user glances from a
menu frame to a text frame.
Linear navigation
If features such as frames and tables are properly labelled
according to the WAI guidelines they should, in theory, be accessible. But
usability may be a different matter. A sighted person can quite easily navigate
within multiple sections of a page in a complex, non-linear or parallel manner.
They might, for example, interact with a search facility to type in terms, or a
drop-down menu to select between choices, while at the same time reading areas
of text promoting specific services. To a non-sighted person, particularly
someone using a screen reader, it may be necessary to navigate the page in a
more serial or linear manner, one frame at a time, with the navigational order
often dictated by the design of the web page, which assistive technology (such
as a screen reader) interprets to read out content in a specific way. This may
result in the need to backtrack a long way (again in a linear manner) in order
to reach the desired point (and then maybe track forward again).
The NoVA project results revealed that, when web pages are
designed to be searched in a non-serial or parallel way (i.e. using frames or
randomly placed links), the navigation process is made considerably longer for
people using screen readers, as they are often forced to search or browse in a
linear or serial way.
The usability tests confirmed that it takes a visually-impaired
user significantly longer to search or browse for information on the web —
particularly if they have little or no sight. This is perhaps not surprising,
but it was also revealed that the time taken for, and the success of, a
searching or browsing task can vary considerably depending on the design of the
site. For example, if good navigational links are provided in an appropriate and
logical order, someone using a screen reader will be taken on a more direct
route to the information they require, rather than having to track back and
forth trying to find relevant information or links. Appropriate use of language
also aids navigation. One user commented: ‘I knew it was a hypertext link, I
just didn’t know where it was taking me to.’
Training issues
The usability tests showed that, although people using the
latest versions of assistive technology were offered navigational shortcuts to
speed up the search and browsing process (e.g. listing hypertext links or
alphabetical sorting), not everyone was aware of these features. Those with more
experience with the assistive technology they were using were often more
successful with the task. It was also noted that some users did not have access
to the latest technologies and so were unable to take advantage of such
shortcuts. Careful consideration must therefore be given to the layout and
navigation of a site and to the different assistive technologies used. Since
success in navigation can also depend on experience with the assistive
technology, there are training issues for both users and trainers.
At the end of each task, users were asked which resource they
would have chosen to look for the information required. Users from both groups
preferred sites with simple, easy-to-use interfaces, that provided reliable
information. Favourites included non-web resources, thus highlighting the
continuing need for a hybrid mix of traditional and digital resources.
The following recommendations from the NoVA project final
report12
aim to address both accessibility and usability issues and to help designers and
providers of web-based resources to move a step further towards universal access
and design:
- Web page design. Take into account not only recommendations of the
W3C/WAI, but also encapsulate user behaviour lessons from the NoVA project and
elsewhere.13
- Assistive technologies. Libraries and museums should invest in
up-to-date technology. However, older versions should not be discarded without
an audit of user requirements in each service. Providers need to be aware of
the trade-off between functionality and familiarity. Web designers must also
be aware of these issues and must not assume that everyone has the latest
version of software available.
- Training. The findings of the NoVA project and parallel research
should be used to inform accessibility training relating to assistive
technology, for example training library and museum staff in relation to open
access computers with assistive technology. Greater emphasis needs to be
placed on training for both users and trainers.
- Universal design. Unless there are pressing reasons not to do so,
all developers should adopt universal design as their underlying aim.
- The appropriateness of digital approaches. In auditing the
accessibility and usability of web pages, service providers should explicitly
consider whether non-web alternatives — including human intermediaries — are
needed.
References
1 J. Owen. ‘Making your website
accessible.’ Library & Information Update, January 2003.
2 J. Craven. Electronic Access for All:
awareness in creating accessible websites for the university library.
University of Dundee: DISinHE, 2000 (www.dmag.org.uk/resources/casestudies/cravenfull.asp#1).
3 L. Hopkins (ed.). Library Services for
Visually-Impaired People: a manual of best practice. Library &
Information Commission Research Report 76. Resource, 2000.
4 www.nlbuk.org/bpm/index.html
5 Royal National Institute for the Blind,
Campaign for Good Web Design (www.rnib.org.uk/digital/welcome.htm)
6 www.whichbook.net/index.jsp
7 www.lhi.org.uk/index.html
8 www.guardian.co.uk/
9 http://www.nlbuk.org/
10 J. Howell. ‘Web
Accessibility and Web Usability: what’s the relationship and why does it
matter?’ RNIB presentation, 5 July 2002 (www.rnib.org.uk/digital/webacc.htm)
11 Web Content Accessibility Guidelines
1.0. World Wide Web Consortium, 1999 (www.w3.org/TR/WCAG10/).
12 J. Craven and P.
Brophy. Non-visual Access to the Digital Library: the use of digital library
interfaces by blind and visually-impaired people. Library & Information
Commission Research Report 145. Centre for Research in Library and Information
Management, 2003. Available to download in Word, PDF or HTML from the project
website (www.cerlim.ac.uk/projects/nova.html).
13 For example, K.
Coyne and J. Nielsen, Beyond ALT Text: making the web easy to use for users
with disabilities. Nielsen Norman Group, 2001.
Jenny Craven is Research Associate, Centre for Research
in Library and Information Management (Cerlim), Manchester Metropolitan
University (mailto:j.craven@mmu.ac.uk.