Skip Nav 

Website test results UK USA and Australia

Page created 19th March 2006.Updated 20th January 2007
Next pages. Design  : Australia  : University  : UK  : USA  : Access  : Guide

Select a page layout and style.
1 :  2 :  3 :  4 :  5 :  6 :  Print :

International web testing results.

Website summary score sheet

Country and webpage
Website Australia UK USA Total Total
W3C Errors Features W3C Errors Features W3C Errors Features
Head FAIL 9 4 FAIL 35 4 FAIL 47 7 91 15
Upper FAIL 10 7 PASS 0 23 FAIL 100 9 110 39
.gov. PASS 0 20 FAIL 46 11 FAIL 51 0 97 31
Minister. FAIL 100 8 PASS 0 18 FAIL 3 10 103 36
Minister. FAIL 4 5 FAIL 28 8 FAIL 23 8 55 21
Minister FAIL 139 3 FAIL 1 11 FAIL 33 8 173 22
Minister PASS 0 18 FAIL 74 10 FAIL 16 7 90 35
E-govt. FAIL 1 10 PASS 0 19 FAIL 19 7 20 36
Human FAIL 11 12 FAIL 1 18 FAIL 84 0 96 30
Surgeons FAIL 13 10 PASS 0 14 FAIL 11 11 24 35
Blind PASS 0 19 FAIL 2 11 FAIL 8 13 10 43
University PASS 0 15 PASS 0 16 FAIL 61 3 61 34
University FAIL 13 8 PASS 0 9 FAIL 100 5 113 22
University FAIL 22 4 FAIL 28 13 FAIL 3 11 53 28
University PASS 0 17 FAIL 17 9 PASS 0 15 17 41
Education FAIL 118 7 FAIL 31 16 FAIL 236 5 385 28
Defence FAIL 1 7 FAIL 1 13 FAIL 39 13 41 33
Courts FAIL 20 12 FAIL 37 6 FAIL 14 10 71 28
Courts FAIL 23 9 PASS 0 18 FAIL 21 4 44 31
Police FAIL 100 8 PASS 0 17 FAIL 66 9 166 34
Police FAIL 5 13 FAIL 19 6 FAIL 71 3 95 22
Post FAIL 154 4 FAIL 100 8 FAIL 185 9 439 21
743 220 420 278 1191 167 2,354 665
Mean 33.7 10 19 12.6 54.1 7.6
∑ ² 79,057 2,758 23,372 4,043 139,501 1,607 1,213
B.S 84,971 596,369

Results

Comparison of means:

Accessibility features and mean error score
Country Pass W3C Pass S.508 Pass AAA Error Access
UK 36% 68% 9% 19 12.6
Australia 22% 59% 9% 33.7 10
USA 4% 36% 0% 54 7.6

Return to top of page

W3C Error validation score:

X = (743 + 420 + 1191)² ÷66 = 83,959.3
B = 552,049 + 176,400 + 1,418,481 ÷ 22 = 97,587.7
BS = 79,057 + 23,372 + 139,501 = 241,930

Accessibility features score:

X = (220 + 278 + 167)² ÷ 66 = 6,700.3
B = 2,758 + 4,043 + 1,607 ÷ 22 = 381.8
BS = 48,400 + 77,284 + 2,582,449 = 2,709,133

Error validation mean scores
33.7 - 19 = 14.7
33.7 - 54.1 = -29.4
19 - 54.1 = -35.1

Accessibility features score
10 - 12.6 = -2.6
10 - 7.6 = 2.4
12.6 - 7.6 - = 5

If F critical < F(K - 1) (N - K) reject Ho that means are the same.
If F critical > F(K - 1) (N - K) accept Ho that means are the same.

Return to top of page

Results Summary

Result: Validation government HTML errors and Country

One of the most obvious results with regard to W3C measured HTML error rates, was the UK government's W3C tested average error rate of 19. The 19 errors per page were partially due to web authoring software. Australian HTML error rates of 33.7 per page are nearly double that number. USA sites on average, had almost three times as many page errors as UK pages. USA errors rates were 54.1 errors per page.

With regard to accessibility features, the UK web pages had the highest and best rate of 12.6 accessibility features per page. Australian feature rates were slightly lower at 10 per page. USA average page features were the lowest at 7.6 per page.

There were many government websites which were XHTML 1.0 Transitional, but only one XHTML 1.0 Strict site which validated, Oxford University.

The House of Lords webpage and some UK government websites were the most accessible government services for the vision impaired, UK sites are generally significantly (.05) far more accessible than Australian or American government websites. UK sites on average, have fewer HTML errors and more accessibility features. As well as the eight UK sites to pass validation, four more UK sites failed validation with only one or two minor errors.

Some UK government websites offer text only versions, different pages for printing, Tab keys for navigation and keyboard shortcuts. Many UK sites also have a link to an accessibility page which lists keyboard shortcuts and has information on browsealoud, a free download to control voices, word pronunciations and speech highlighting.

In Australia, the special minister of State Gary Nairn has stated that Australia is a leader in e-government, but there were only five web sites which validated as error free, but two of these were quasi-governmental health and educational concerns, Vision Australia and Melbourne University.

Return to top of page

Discussion

Karl Popper said that in trying to show that all ravens are black, one white raven ruins your hypothesis. BUT THIS IS NOT SOCIAL RESEARCH. The study used a W3C measure of error rates, an objective accessibility online validation and measure of HTML code accessibility features. These are easily measured variables, compared to social research data about real people with visual impairment! The assumption is that the sample and study design tried to compensate for not having the resources to test real visually impaired people and get their opinion. In theory, I try to cater for the colour blind on this site. In theory, but I would really appreciate feedback from real world people.

There were the occasional white ravens in this study. Victorian State Premier Brack's website and a few other web pages which did not W3C validate, but were proirity one accessible, mainly American sites, in compliance with the American S.508 criteria.

In this small sample, the American page validation rate was only 4% (Princeston University). Previous larger sample research by Krantz 2005 found that only 14 of 546 (2.5%) of American government websites validated as HTML documents. Seven of the American pages tested passed basic accessibility S.508 tests, but were invalid as HTML documents. There is little integration in the measures. Accessibility tests for a list of attributes or absence of faults, but HTML document errors could cause the page to not be displayed at all.

Return to top of page

CSS Stylesheets

There was hardly an innovation in the use of stylesheets, most sites only had one or two stylesheets. The Royal college of Surgeons provided larger print and projector stylesheets.

Stylesheets can not only control fonts and formatting, they can add text or numbers that do not exist in the HTML.

For one example of using CSS, in Patrick Hanrahan's novel NUNC Set the stylesheet to Number two The author's choice. Stylesheet 2 uses the before: property to insert the characters names before their lines in a white #FFFFFF font on the black background. Patrick's multilingual novels could be read by multilingual voice technology, language HTML font tags are included for each character's lines, Latin, Italian, French, Spanish, Greek, German or Russian.

Each character in the novel has their own font and colour. They could also have their own voice accents.
.LeadMale:before

{content:"LeadMale ";font-size:90%;font-style:italic;
color:#FFFFFF;voice-family:male;voice-family:Ralph;
azimuth:center;elevation:15deg;pause-before:1s;
speech-rate:medium;pitch:low;angle:center-right}

Stylesheets can do more than just format layout fonts and text, they can add extra interactivity and information before or after or they can even insert content in any HTML page element.

Return to top of page

Statistics of visual impairment

Accessibility and colour blindness
1.5 million Americans are visually impaired. Over 8% of adult males but less than 1% of adult females have some form of colour blindness.

Legal considerations

The extent to which government websites make the effort to produce validated and accessible webpage may be related to the legal requirements to do so. For example, USA webpages were on average the worst we tested. This may be because they aim for lower standards, A one S.508. If it passes the lowest accessibility test it is acceptable.

American Law

Section 508 of the U.S. 1974 Rehabilitation Act prohibits federal agencies from buying, developing, maintaining, or using electronic and information technology that are inaccessible to people with disabilities. As of June 21, 2001 all federal agencies were mandated to have their web sites compliant to section 508. The regulation is available at:
Us Act Formal online validation and testing of these rules Criteria. See Cynthia validator.
In the case of Gumson, a blind American used a screen reader to speak web page text to him. He was trying to buy an airline ticket online. As in th example above, because of missing alt tags, tables with no headings, and the lack of keyboard or tab index page navigation, it was impossible for Robert Gumson to buy a Southwest Airlines ticket.

A Mid level American Circuit Court of Appeal, ruled Gumson had no right to sue the airline's website under the Americans with Disability Act (ADA). In 2002 it was ruled that ADA did not apply, as the "internet was not a physical, concrete places of public accommodation". Gumson appealed. Later that year in 2002, a higher court, the Florida Federal court heard the case on appeal. They again ruled for the airline and upheld the ratio of the lower court.

Target in America.

UK

UK government regulations
Some advice on the Guidelines for the UK
UK Disability Discrimination Act 2004.

Australian

Disability Discrimination Act 1992 Section 24
Unlike USA court rulings on Section 508, that the internet is not a real place, the Australian legislation applies to the internet, government agencies and private companies. Section 24 of the Act ("services and facilities") states:

1. It is unlawful for a person who, whether for payment or not, provides goods or services, or makes facilities available, to discriminate against another person on the ground of the other person's disability or a disability of any of that other person's associates:
1. by refusing to provide the other person with those goods or services or to make those facilities available to the other person; or
2. in the terms or conditions on which the first-mentioned person provides the other person with those goods or services or makes those facilities available to the other person; or
3. in the manner in which the first-mentioned person provides the other person with those goods or services or makes those facilities available to the other person.
2. This section does not render it unlawful to discriminate against a person on the ground of the person's disability if the provision of the goods or services, or making facilities available, would impose unjustifiable hardship on the person who provides the goods or services or makes the facilities available.

Responses to HTML accessibility whistleblowers

I decided that this study was a wasted effort, if I did not email government departments to inform them of their HTML errors and missing accessibility features. When I emailed the Australian Finance department and AGIMO about errors in their pages, they replied "We lead by example" I do not believe that attitude is conducive to learning what is HTML best practice! Perhaps it is even, a parochial Australian attitude? I emailed the UK government who are updating their websites all the time. No wonder English sites are better. I believe that English advertising for employment in website development also show more accessibility emphasis than Australian or American positions advertised.

I received a much more; considerate, best practice, W3C validated, AAA accessibility reply from the UK cabinet-office on April 4th 2006.

Tim,

Many thanks for your kind words. We are in the process of migrating all of the sites that we work on onto a new system and as they go live on the new platform we are trying to ensure they all pass as valid XHTML and AAA accessible. If you would like I can let you know what sites are migrating and let you know once they are live - including www.cabinetoffice.gov.uk and www.civilservice.gov.uk.

Already live are www.lordsappointments.gov.uk as you are aware and www.acoba.gov.uk which you may or may not be aware of.

I had time to have a quick look at the stats you are working on and look forward to seeing the finalised version.

Kind regards
cabinet-office.x.gsi.gov.uk

Australia's Prime Minister John Howard's office replied, pretending that he is attending to their accessibility electoral duties for e-governance as well as their HTML errors. Unfortunately Mr Howard's site has not improved at all since he was first notified of the HTML errors in his homepage. Nine months is a long time to do nothing! I hope the consultant was not paid much, they have done nothing!

I had not checked www.acoba.gov.uk Another great UK site, keyboard shortcuts, tab key page navigation. I missed some UK sites which have been updated. The UK government is progressing very quickly, not by writing or talking about "better practice", they are actually implementing best practice. They even respond politely to a humble antipodean member of the Commonwealth. I did not even bother to email any American government websites.

It would be more reliable to take a larger stratified random sample of 50 government webpages from the Heads of State down to local government. Anyone able to provide funding for a larger study would be welcomed. 22 sites from three different countries was all I could sample. I would like to have sampled 100 sites from each country and to at least have included Canada and Australia's close neighbour New Zealand.

Many other countries could be included in a world wide study. Both the sample size and diversity of countries sampled could be improved. There may be some fantastic country out there doing a great job with validity and accessibility that I am unaware of?

The rated 24 accessibility items were all equally weighted as being worth one point, it might be that for a blind user, keyboard shortcuts and tab key page navigation is worth much more than a copyright meta tag stating who wrote the page and when. These 24 accessibility features could be tested first, asking vision impaired users to rate them in importance. A weighted scale could then be used to attribute values to each feature, depending on how critical it is.

Return to top of page

Conclusion

Some people argue that if a web page looks OK, then everything is good. But if you want a Rolls Royce on the internet, one that any citizen can access through a keyboard, you help provide the best possible e-government service.

As well as equity considerations and relevant laws in each country, there are also search engine optimisation (SEO) reasons why you should provide valid accessible web pages. The only disadvantage of success is that search engine web-bots will devour with relish your validated W3C HTML and strip every graphic from your site. Fend off email collectors and bad-spam-bots from Beijing almost every-night.

In conclusion, the only country of the three tested, who can claim to lead in e-government by example, is the UK.

Proposals

Please contact the Webmaster, if you want to monitor your organisation's websites for W3C compliance and as a legal prophylactic, demonstrating compliance with the highest international disability access standards. Contact Heretic Press to set up an accessibility audit of three pages of your website. Initial charges are USD $500 for an accessibility audit of your website.

Accessibility reports

Instead of procrastinating like Target, start now to build a better website, increase your web traffic and legal compliance. For USD $500 we can provide you with a comprehensive accessibility report on your homepage's compliance with UK, USA or Australian laws. The report will include results from automated testing tools and an analysis of the code used in your webpage. The report will identify where your website should be improved to meet the highest international standards of AAA web accessibility. Reports are completed within 48 hours of payment being received. Please contact the Webmaster if you have any special requirements.

Payment for Accessibility Report

Why expose your government or organisation to claims of disability discrimination like Target in America. Instead of taking years to do something about accessibility, your accessibility report from Heretic Press will generally be completed within 48 hours of payment being received.

Heretic Press offers you many options to check the progress of your website, making it easy to monitor compliance with international accessibility best practice by:

1. Listing your page elements in an easy table checklist.
2. Providing W3C compliance advice and HTML code suggestions.
3. Detailing accessibility features required.
4. Specifying areas to improve and template design.

Heretic Press

Seamless looping earth and satellites gif animation preview only Heretic Press phone animation page.

hereticpress.com
W3C test Heretic Press

Verified File Name: Heretic Press
Date and Time: 12th December 2006 23:27:
Result: PASSED validation
No HTML errors

Cynthia Says® - Web Accessibility Report
Verified File Name: Results.html
Date and Time: 20 December 2006
PASSED Accessibility Verification priority 1, 2 and 3.

Validation Error Score Results.html
No Errors
Page Accessibility features Results.html 24

Validity Results.html 20th December 2006

Variable Result Variable Result
W3C Errors No Errors Links PASS
CHARSET ISO-8859-1 HTML version XHTML 1.0 Strict
Page Language English HTML color PASS
NOSCRIPT PASS HTML link text PASS
image borders PASS Rollovers onfocus
Stylesheets PASS Meta Copyright Heretic Press
Meta modified Tue, 27 Mar 2006 Meta parents children Siblings
Cynthia™ 508 PASS Large Text PASS
Cynthia™ One PASS ALT tags PASS
Cynthia™ Two PASS LONGDESC LONGDESC=
Cynthia™ Three PASS Skip navigation PASS
Tab index TAB KEYS Keyboard Shortcuts PASS
Page features 24
Validated XHTML 1.0 Strict W3C small logo :  Valid CSS! :  W3C-WAI Web Content Accessibility Guidelines :  Cynthia Tested! :  Business publishing Excellent accessibility rating from net progress : 

Contact Heretic Press for an accessibility review of your website. Our editor is recognised internationally for accessible online publishing and he is an experienced educational publisher of science history and psychology textbooks and needs the work. Audits will help you identify current accessibility problems and provide guidance for compliance with international standards and best practice.

Heretic Press Accessibility Statement
Heretic Press Guide to writing HTML
Contact Heretic Press

Return to top of page

Other Accessibility research

Canadian site study
Dutch accessibility study
50 US sites Neither the mandate nor the motivation.
eGovernment Site Credibility: Comparing Speed, Accessibility, Typos, and Validity eGovernment Site Credibility
Web Page Accessibility: Research Studies
Web Accessibility Survey Site

Web publishing guidelines

UN Report
UK Web Guidelines

Legal Obligation Liability

UK legal firm
Website sued

Return to top of page