The DRC report into UK web accessibilityWednesday, April 14, 2004
The DRC releases its report on the accessibility of UK websites today. An inkling of what to expect is already available through a news article on ePolitix titled: "Website access 'denied to disabled people'". The first statistics are:
the Disability Rights Commission found that, of the 1,000 public websites examined, 81 per cent failed to meet the minimum standards for disabled web access.
The DRC have released a press statement already, and Bert Massie, DRC Chairman continues:
"The situation revealed by this investigation is unacceptable but not inevitable. The DRC is determined to ensure that this new powerful technology does not leave disabled people behind."
My intention is to update this particular blog entry as more details emerge.
Notes from the DRC web cast:
Helen Petrie: about the survey
A panel of 50 users was formed, taking people from a wide range of disabilities, ages, ethnicity using a broad range of assistive technologies. They first conducted interview, focus groups to identify typical good and bad website and reasons why they were good or bad.
The representative sample took 1000 websites covering the related interests of disabled people. This covered the following areas:
- Government and information websites
- E-commerce sites, for example banking, travel and retail websites
- Entertainment and leisure websites
- Web services, for example ISPs, portals and search engines
First stage of testing
The first stage was the automated testing of each website's home page. The automated testing was done using commercial software using WCAG as a guideline. Using commercial tools, only 13 of the 56 checkpoints could be tested automatically.
The main findings were that only 19% of websites passed level A compliance - that is the most basic of accessibility levels. The interesting point in this section was that Government and information sites, 30% of these sites passed the basic accessibility level.
Priority 2 compliance. The UK Government encourages its web developers to meet level 2 compliance. But only 6 websites of the 1000 passed the automated level 2 checks. With a proper level 2 compliance against the WCAG only two of those sites complied.
None of the 1000 websites tested passed priority 3 automated tests. The conclusion is very poor.
Second stage of testing
From the original batch of 1000 websites, 100 were selected for a more detailed study. The sites selected included all five categories of websites plus sites using technology such as Flash and tables. An automated test was run on the entire website - or the first 500 pages if the was that big. As a whole 39,000 web pages were tested, making this the largest and most comprehensive study of web accessibility in the world.
The panel, accompanied by researchers, went into the DRC lab and explored a website. Two representative tasks were assigned for each website. (For example, on a banking site, the task was to find the current interest rate). They were then asked to complete a questionnaire rating various characteristics of the website as well as answering open-ended questions about problems experienced on the website. Then to continue, the panel were asked to do this test again at home or in the office with seven or eight other websites. Each member of the 50 person panel checked 10 websites. 1000 tasks were created allowing two tasks per user per website. A total of 913 tasks were completed.
The results were surprising. Only 76% of the basic tasks were completed (the expectation was higher for such simplified tasks). Blind users only successfully completed 53% of the tasks. Partially sighted people were also disadvantaged and only experienced a 76% success rate. Dyslexics, hearing-impaired, physically impaired users scored an 80% success on these simplified tasks.
Five hundred and eighty five problems were encountered. 55% of these were covered - and could be resolved by following the WCAG. 45% of problems didn't relate to the guidelines. The conclusion was that web developers absolutely need to follow the Web Content Accessibility Guidelines, but also to have usability testing with disabled people. The guidelines are necessary, but not sufficient.
Of the 55% of problems covered in the WCAG, 82% of them could be resolved by following eight specific checkpoints. If web developers addressed those eight checkpoints, the web would not be perfect, but it would be more accessible and usable to people with disabilities. These eight checkpoints should form the bedrock of compliance.
Accessibility compliance does not hinder creativity or innovation. These eight checkpoints are simple to meet.
Usability bonus of accessibility
Part of the research involved a controlled study taking 6 websites - three with high accessibility and three with low accessibility to determine the disenfranchisement of non-blind users. For a high accessible website it took 36 seconds for a non-disabled person to complete the tasks. On a low accessible website it took non-disabled people 52 seconds to complete the tasks. The conclusion is that high accessibility improved the usability of websites for the non-disabled audience too.
Extreme website testing involves following the WCAG guidelines as well as conducting usability testing with disabled people. This form of testing addresses both accessibility and usability. This is the usability bonus.
The recurrent barriers to access
Seven hundred website owners and web developers in the public and private sector were asked to complete an accessibility survey. Only 89 replies were received - this may be a result of apathy or lack of knowledge around accessibility. Twenty-one website owners and twenty five web developers - randomly selected - were interviewed to supplement this survey.
Results from website owners were that in large organisations there was a high degree of awareness of web accessibility and a good understanding of the responsibilities of the Disability Discrimination Act. This is surprising in contrast by the very low numbers of accessible sites. Statistics from small businesses are less positive, they are less aware of the issues and responsibilities of web accessibility and the DDA. Only 29% of them took accessibility into consideration in designing and building their websites.
Website owners perceive the following problems with accessibility:
- They were worried about the costs of implementing accessible guidelines
- They lacked the knowledge about the issues and responsibilities of web accessibility
- They believe web accessible design is incompatible with sophisticated websites
- They believe web accessibility is incompatible with creativity and innovation
The last two points were highlighted by Helen Petrie as being myths.
Eighty percent of web developers indicated they had attempted accessible development sometimes. They indicated that it was the lack of interest from clients about accessibility that was the reason they produced inaccessible websites (quite a contrast from what the clients indicated!). Web developers felt the best argument for creating accessible websites was that it increased the potential audience for that website. They admitted they lacked the knowledge and experience for building accessible websites, only 9% claimed any experience in accessibility.
The conclusion was that accessibility in Great Britain is very poor. The blind are the most disenfranchised from it, followed by the partially sighted and dyslexic audience. Web sites that meet only eight of the WCAG checkpoints will make a lot of difference. Automated testing is important, but not sufficient to determine whether a website is usable to people with disabilities. User involvement in testing is very important in addressing accessibility and usability. Awareness and training is vital and important for both website owners and web developers.
Michael Burton: the DRC Recommendations
The situation about web accessibility is completely unacceptable. The task failure rate of over 50% is appallingly high. Disabled people already suffer from a depressed productivity rate, and badly designed website further impacts productivity by another 50%.
Many disabled people are experiencing less favourable treatment with UK websites. It is unethical, inconsiderate and ill informed as well an an unnecessary penalty to the disabled. It is uncommercial - an accessible website benefits disabled people as well as improving usability for non-disabled people. The DDA requires sites to be acceptable to disabled people, so these websites are illegal.
The principal recommendations cover four groups of people:
Ultimate responsibility for accessibility rests with the website owners. Accessibility issues are not high on their agendas. The awareness of the issues and responsibilities is claimed, but it is not translated into effective action. Web designers also report clients are not terribly interested in the subject.
Board and governing bodies of companies and organisations must adopt formal policies for meeting the needs of disabled people. They need to demonstrate reasonable adjustments and must show consideration to disabled needs. Policies must cover issues such as the level of accessibility required. Specifications and invitations to tender must document include compliance to this policy. This is a compliance issue - like Health and Safety in the Workplace Act, the Data Protection Act, and Financial Services Authority compliance. Boards and governing bodies can delegate responsibility for accessibility, but cannot abdicate responsibility. The suggested course is to assign responsibility and then audit for compliance.
The DRC report serves notice with this report that the DRC will use all its powers to seek compliance on this matter. The DRC has legal power and is entitled to start legal proceedings where necessary.
Website owners need to rely of web developers to interpret and implement accessibility policy. The web design industry is both fragmented and immature, populated by creative and independent people. There is no established and professional structure in place.
Web developers need to better understand the needs of disabled people. They need to acquire the skills to respond to these challenges - for instance knowing what clear and consistent navigation is. Many accessibility issues are practical issues. A web developers understanding most not rely completely on web standards and tools - they don't guarantee accessibility. They need to include disabled people in design and testing.
How can web developers acquire the necessary skills? The recommendation is on educators and trainers to include accessibility in their material. Include modules for disability awareness and accessibility. The DRC or the government needs to contact vendors of authoring tools as well as tackling 1 to 7 day courses on website development - the bedrock of training in this field. The DRC does not have the resources or skills to tackle these problems.
Good website design achieves very little if disabled people are not equipped with the right hardware, software and training. These are all hurdles to disabled people. Some organisations select the wrong tools and lack the training to utilise them correctly.
There are very limited number of good sources of advice, and they mostly reside in charity organisations and voluntary support organisations. They also have limited capacity - Abilitynet for example has a limited capacity of 2000 assessments a year.
The UK Government recognises the web is important and has expressed concerns about the digital divide. The danger is that disabled people will be on the wrong side of this divide. Government has a responsibility for bridging this gap. The summary of government efforts so far is: "They could do better."
The DRC believes that government has an important role as a catalyst in promoting professionalism (in the web development industry). The Government are advised to support public campaigns to website owners about the needs of disabled people and their duties under the DDA. The government needs to foster professionalism, and there is a desperate need for best practice guidelines. The Government needs to take the initiative on this matter (they may not be qualified to write the guidelines, but they can certainly bring together the right people).
Collectively, users have a wealth of knowledge and experience. This is not an easy resource to tap. Many skills needed are disability specific. Organisations should offer disability specific advice and experience enabling them to share this knowledge amongst their users.
Vendors are the source of testing tools, development tools and generic assistive technology. The solution lies in raising awareness and changing perceptions. The DRC report is hopefully instrumental in convincing vendors for the need to improve their products for accessibility.
The main drive can only come from people who commission and pay for websites. They must step up to their social, legal and commercial responsibilities, and must view accessibility issues as a fundamental part of their business case for investment.
Question and Answer session
What is the next step? The immediate task is to identify partners in business and industry to champion the recommendations. The DRC needs to find suitable partners in the web development industry and network out from there. The web development industry themselves need to champion the issues.
Will it require a legal case before people will obey the law? The principle driver is the website owner - he has the legal obligation to make his website accessible. Court cases will be a wake-up call. The website owner is liable in law, not the webdesigners. The DRC approach is to work with the industry. With 1000 websites, not a single one met level AAA compliance - that means almost every website in this country can be liable.
When a disabled person calls the DRC complaining about a website they can't access, the first step is to call the owner of that website - that normally is sufficient to resolve the issues. The DRC does have legal powers and will use them. The DRC wants to work with you. The research is robust, the conclusions are clear, and there are answers for improvement. The DRC is using the force of argument. If this fails, the DRC will not hesitate to use legal force.
Were the websites tested told? The DRC do not know which websites were tested. The purpose of the survey was to produce statistically valid conclusions, not to identify good or bad websites. The identity of the websites were concealed from the DRC.
When it is reasonable to comprimise WAI for style? It is not necessary since there isn't a compromise between style and accessibility.
Won't rubbishing the WAI standards cause a breakdown? It is not the DRC's intention of rubbishing the WAI. Conformance with the guidelines alone is not sufficient to the objective of access to disabled people. These guidelines need to be complemented.
Does the DRC take a stance on 2v1 site issue? (Having an accessible version of the website alongside an inaccessible one) Tesco & Manchester United have separate accessible versions of their websites for the disabled. Helen Petrie: We appreciate web developers making the effort to produce two sites. Not offering the same functionality on both is not acceptable. Creating two sites is not a necessary effort. Web developers using tools imaginatively and creatively can create sites accessible and usable to everyone. The power of the web is providing information usable by all users. We discourage developers from creating two websites.
If the only way is two sites (as a last resort) - then okay. But being inclusive is the ideal. Improving accessibility improves the website for all users - as the report proves. The one caveat for the two site approach is when Flash is being used. Flash is unlikely to be accessible, so provide an alternative way of accessing that material.
Any response from the government? Today is the launch. The DRC will be communicating with the DTI - Patricia Hewitt.
What will it take for accessible websites to be the norm? Is a high profile website court case essential? The DRC has brough high-profile cases (Ryan Air and schools). No doubt that publicity does focus people's minds. (To the reporters): Write your articles well then we may not need a court case!
Part of the strategy in funding cases is to find the right sort of case - one that can clarify obscurities. Settlements are one avenue service providers take since it is more cost-effective to make changes to the website than a court appearance. The question is of picking the right case - something like the Australian Olympic case.
Julie Howell, from the RNIB: The RNIB endorses cases. The two RNIB cases last July were both settles. The settlement was in the best interests to the blind and partially sighted people. The RNIB welcomes the report, its findings and recommendations. What about highly transactional websites?
Helen Petrie: Highly transactional websites fared no worse than other areas. Although this means that they were not very good. Web developers will require specific guidelines for different areas - for example transactional sites. They struggle with completing forms. A code of practive for web developers will be very useful.
The headlines are:
- Many businesses are not conforming
- Most businesses are vulnerable to a legal challenge
- 81% of websites fail minimum levels of access
- The majority of problems were not contained in the WAI guidelines (WAI dispute this)
- The knowledge deficit is so severe that government must consider involvement in promoting accreditation and registration (presumably of web developers)
Online coverage or commentary
- BBC: Websites 'failing' disabled users
- WAI: Statement on web access report from UK DRC
- NetImperative: UK fails disabled internet access test
- out-law: Disabled web access being ignored, reports DRC
- Internet Magazine: Accessibility study calls for government intervention and asks for WAI update
- vnunet: Sites face legal action on disabled access
- BBC: Websites 'failing' disabled users
- Juicy Studio: Formal investigation of UK websites
- Computer Weekly: Ignoring disabled web access will lead to legal action, warns DRC
- out-law: Egg decorated for site accessibility in DRC study
- BBC: A web open to all?