Difference between revisions of "User:Sweerdenburg/BTH740 ResearchEssay"

From CDOT Wiki
Jump to: navigation, search
(Research Notes)
Line 7: Line 7:
==== Thesis ====
==== Thesis ====
With the internet becoming increasingly technologically diverse and pervasive to our daily lives, just how inclusive is this environment becoming for those with special needs?
With the internet becoming increasingly diverse and pervasive to our daily lives, how well are standards and accessibility able to keep up with technology and each other?
<!-- With web technologies evolving to allow more creative and striking web displays on computers, how are developers addressing accessibility of less capable devices such as cell phones and PDAs? -->
<!-- With web technologies evolving to allow more creative and striking web displays on computers, how are developers addressing accessibility of less capable devices such as cell phones and PDAs? -->

Revision as of 22:07, 7 November 2010

BTH740 | Weekly Schedule | Research Projects | Research Essay | Student Resources

WCAG on the Modern Web

Steven Weerdenburg Email

Thesis Statement


With the internet becoming increasingly diverse and pervasive to our daily lives, how well are standards and accessibility able to keep up with technology and each other?


Accessibility, Evaluation, Education, Section 508, people with disabilities, guidelines, methodologies, metadata, contextual design, automated usability evaluation, design guidelines, empirical studies, usability, web site design, ATAG, UAAG, W3C, WAI, WCAG, World Wide Web, Internet, ageing, disability, impairment, older people, usability, device-tailored evaluations, mobile web, web device, disability rights


  • S. K. Kane, J. A. Shulman, T. J. Shockley, and R. E. Ladner. “A web accessibility report card for top international university web sites.” Proceedings of the 2007 international cross-disciplinary conference on Web accessibility (W4A) 225 (2007), 148-156. ACM. 17 October 2010, <http://portal.acm.org/citation.cfm?doid=1243441.1243472>
  • D. Sloan, A. Heath, F. Hamilton, B. Kelly, H. Petrie, and L. Phipps. “Contextual Web Accessibility – Maximizing the benefit of Accessibility Guidelines.” Proceedings of the 2006 International Cross-Disciplinary Workshop on Web Accessibility (W4A) 134 (2006), 121–131. ACM. 17 October 2010, <http://portal.acm.org /citation.cfm?doid=1133219.1133242>
  • M. Vigo, A. Aizpruna, M. Arrue, J. Abascal. “Evaluating web accessibility for specific mobile devices.” Proceedings of the 2008 International Cross-Disciplinary Workshop on Web Accessibility (W4A) 317 (2008), 65–72. ACM. <http://portal.acm.org/citation.cfm?doid=1368044.1368059>
  • L. G. Reid and A. Snow-Weaver. “WCAG 2.0: a web accessibility standard for the evolving web.” Proceedings of the 2008 International Cross-Disciplinary Workshop on Web Accessibility (W4A) 317 (2008), 109-115. ACM. Seneca College of Applied Arts and Technology, Toronto, ON. <http://portal.acm.org/citation.cfm?doid=1368044.1368069>
  • J. Clarkson, P. Langdon and P. Robinson. Designing Accessible Technology, 199-208. London: Springer (2006).
  • S. Horton. Access by Design : a Guide to Universal Usability for Web Designers. Berkeley, CA: New Riders (2006).
  • J. Sydik. Design Accessible Web Sites: Thirty-six Keys to Creating Content for All Audiences and Platforms. Raleigh, N.C. : Pragmatic Bookshelf 2007

Research Notes

Preliminary notes, Non-prioritized into an argument PDF (includes hand-written)

Finalized PDF

-- Get into overview of evolution of web technologies. HTML 4, HTML5, CSS, JavaScript. Include user-supplied content (wikis) and the like (Flickr, social media).

Concern about web site accessibility has been growing steadily since the mid 90s (Kane, pg 2).

WCAG 1 was for static web sites, now dynamic sites are much more common. Much more involving technologies. WCAG 1 was HTML specific though WCAG 2 is technology neutral. Subjective definitions of WCAG 1 posed issues for testability (Reid pg 109).

4 Criteria (Reid 110, Sydik): Presentable Observable Understandable Robustness

3 levels: A, AA, AAA based on possibility, applicability, reasonability, limitations imposed and existence of workarounds. Levels build on each other (A must be met to be AA, AA must be met to be AAA). 65 total checkpoints (Reid pg 110).

Despite a large amount of work into legislation, authoring tools, assistive technology development and best practice identification. Large lack of awareness despite efforts of the Web Accessibility Initiative (WAI). WAI developed Web Content Accessibility Guidelines (WCAG) and the User Agent Accessibility Guidelines (UAAG) and Authoring Tools Accessibility Guidelines (ATAG). (Sloan, pg 122)

Standard-adherence is important to screen readers as pointed out by Horton, Reid.

Standard adoption is similar to industrial revolution standardization of nail/screw sizes, though difficulty is analogous to having living-room plays, gym recitals, small theatre productions and grand operas all appear one after another on same stage (Kreps, pg 2). Mentions though that the artisans who yielded control in Industrial gave up easier than Microsoft has (Kreps, pg 3), and that W3C publishes "recommendations" rather than "standards".

WCAG took nearly 8 years to come out • Initial Draft Jan 25 2001 - http://www.w3.org/TR/2001/WD-WCAG20-20010125/ • Final Draft Dec 12 2008 - http://www.w3.org/TR/2008/REC-WCAG20-20081211/

(Reid, pg 113) Accessible Rich Internet Application (ARIA) • Modernized, Extension of XHTML • Defines mechanism for mapping custom UI controls to native OS accessibility frameworks • Defines "role" attribute for all elements • "State" and "Property" attributes that can be mapped to accessibility APIS (onMenuOpen, etc) • Keyboard focus (extends tabindex) • Notification of changes by defining events • Limited support (NEED UP TO DATE INFO, ONLY LISTS FF3, "upcoming IE8") ○ IE8 more standards compliant and accessible http://blogs.msdn.com/b/ie/archive/2009/01/16/accessibility-improved-aria-support-in-the-ie8-rc.aspx ○ GoogleReader http://googleblog.blogspot.com/2008/03/aria-for-google-reader-in-praise-of.html ○ Firefox 3.5, 4 ongoing as standard changes § http://www.marcozehe.de/2009/06/26/new-accessibility-features-in-firefox-3-5/ § http://www.marcozehe.de/2010/10/04/new-in-accessibility-in-firefox-4-0/ ○ Webkit (Safari, Chrome) http://www.paciellogroup.com/blog/?p=454, http://www.paciellogroup.com/blog/aria-tests/ARIA-Safari-RoleTests280909.html

ARIA STANDARDS CHANGED OVER TIME WHILE FINALIZING. FF CHANGED WITH STANDARDS, WEBKIT ADOPTED LATE, IE ADOPTED MIDDLE. STILL NOT FINALIZED (http://www.w3.org/TR/2008/WD-wai-aria-20080204/ = Initial, http://www.w3.org/TR/2010/WD-wai-aria-20100916/ = Most recent).

Above Aria adoption is not alone, HTML5 adoption is similarly sporadic across browsers • IE has no html5 until IE9 (http://msdn.microsoft.com/en-us/ie/ff468705.aspx) • Firefox ○ Had canvas since 2005 (Gecko 1.8 aka FF1.5) https://developer.mozilla.org/en/Canvas_tutorial) ○ Had video since 2007 (Gecko 1.9.1 aka FF 3.5) https://developer.mozilla.org/En/Using_audio_and_video_in_Firefox

Factor in browser usability statistics (Kreps, pg 4), but NEEDS UP TO DATE INFO, DATED 2008 (All below from http://en.wikipedia.org/wiki/Usage_share_of_web_browsers) • NetApplication (40,000 sites, unique visitors) http://marketshare.hitslink.com/browser-market-share.aspx?qprid=0 • W3Counter (last 15,000 non-unique views) http://www.w3counter.com/globalstats.php • StatCounter (3,000,000 sites, #hits (not unique)) http://gs.statcounter.com/ ○ Can drill by region/country. Huge differences between US and EU • W3Schools Counter http://www.w3schools.com/browsers/browsers_stats.asp • StatOwl (USA only) http://www.statowl.com/web_browser_market_share.php • Wikimedia visitors http://en.wikipedia.org/wiki/Usage_share_of_web_browsers#Wikimedia_.28April_2009_to_present.29

"Indeed their victory over Microsoft and Netscape in the Browser Wars of the mid-late 1990s was something achieved rather through the vigorous lobbying of external organisations such as the Web Standards Project, (WaSP 1998) formed to enable web developers to avoid the increasingly necessary expense of creating multiple versions of their websites individually tailored to increasingly different browsers." (Kreps, 3). Almost presents same issue again that occurred in mid 90s which caused W3C to define standards/recommendations in place of IE and Netscape, except coding for various implementations of client technologies (goes back to User Agent Accessibility Guidelines). WHAT GOOD IS TECHNOLOGY IF CAN'T RELY ON IT'S USAGE? WHY PUT EFFORT INTO DEVELOPMENT?

Development of a product is further refined as recommendations from the W3C or other bodies are adopted into legislation by state bodies. "WCAG has also been used as a model for accessibility legislation, such as in the Section 508 guidelines for US government web sites [34]." (Kane, 149)

Numerous State bodies have passed binding legislation with regards to WCAG (Sydik, 292-297) State Body Legislation Scope Information Australia Disability Discrimintation Act Public/Private Sector http://hreoc.gov.au/disability_rights/dda_guide.htm Canada Canadian Human Rights Act, Ontario Human Rights Code.

All sites with WCAG 2 AA Accepted to be Private (Human Rights)

Public (WCAG) http://www.tbs-sct.gc.ca/clf2-nsi2/

http://www.zvulony.com/accessibility.html EU eInclusion None (many member states have legislation for public/private sector) http://ec.europa.eu/information_society/activities/einclusion Japan Japan Industrial Standard (JIS) X 8341-3 Non-legal, considered best practice http://www.mitsue.co.jp/english/column/backnum/20040625.html UK Disability Discrimination Act

Special Educational Needs and Disability Act Public/Private

Education http://www.drc-gb.org/PDF/CoP_Access.pdf

(Kane, 149) UN Convention of Rights of Persons with Disabilities 100 member nations http://www.un.org/esa/socdev/enable/

http://www.nomensa.com/resources/research/united-nations-global-audit-of-accessibility.html US Section 508 Public http://www.section508.gov/

Further Law Info can be found at WebAIM World Laws (http://www.webaim.org/articles/laws/world/)


Dexter Hypertext Model Overview (Dodd, pg 4) Document may be atom (single page) of Composite component. Anchor - Link between documents. Can be scriptable, determinable at run time (issue for screen readers?)

The Dexter Model is formally expressed as three distinct layers as shown in Figure 2 below; the relevant items from Figure 2 are listed for each layer. (Dodd, pg 5)

Figure 2 – The three layer Dexter Model

"Each Component within the Dexter Model is associated with a Presentation Specification. The Presentation Specification describes how an individual component should be rendered to the user." (Dodd, pg 5)

Design-wise this is accomplished through CSS, but User Agent-wise this means updating CSS interpretations within the browser and releasing new browser versions. This is reliant on standards.

Many automated tools exist to test, though false positives are potential. WebXACT (aka Bobby) and Cynthia Says are two common WCAG 1.0 testers. FAE tests modification of this to both WCAG and Section 508 (US only) legislations. Alt Text on images is commonly flagged, but is alt text significant on space-holding images? WebInSight tries to mitigate this error. (Kane, 151).

Of the 778 sites tested by Kreps, a vast majority had many many issues. Were outright invalid or were not up to standard (Kreps, 6).

REFERENCES TO UNIVERSITY WEB SITE STUDIES (Kane , 149). <25% of university home pages were passed by Bobby (Kane from Rowland, 149) 77% of home pages or pages linked from were failed by Bobby from to p24 ranked info science universities in US (Kane from Schmetzke, 149). Automated tools to test top 50 US schools, 15% failed in at least accessibility or usability tests, 35/50 failed WCAG Priority 1 (Kane from Zaphiris and Ellis, 149)

Kane conducted self study on Times Top 100 Universities using multi-method technique (Bobby and Cynthia scores combined, FAE kept separate). Only 2 were 100% WCAG 1 compliant (Bobby and Cynthia), while "Among those countries with 5 or more ranked universities, universities in Australia contain the fewest average accessibility errors, followed by the United Kingdom, the United States, the Netherlands and France (tie), and Switzerland." (Kane, 152). Results found to be statistically significantly. Table on 152 shows numbers, DO THESE COINCIDE WITH SCOPE OF LEGISLATION (PUBLIC/PRIVATE VS. PRIVATE ONLY)? FOR THOSE THAT DO, DOES RATE OF ADHERENCE COINCIDE WITH DURATION OF TIME SINCE LEGISLATION ENACTED (LESS INFRACTIONS FOR LONGER TIME ENACTED)?

Aside from markup itself, also worth noting is the additional implementation time of drafting and updating accessibility policies. Kane found less than half (46/100) sites had accessibility policies somewhere on the site, with a fifth (21/100) linked from the home page. More overhead in adherence than simply markup, but promotes best practice: "Web sites that featured accessibility policies had statistically fewer accessibility violations than sites that did not have policies (p < .01) Sites without policies averaged 5.36 P1-P3 violations, while sites with policies averaged 3.98 P1-3 violations." (Kane, 154)

"The issue of web accessibility has received a fair amount of attention in English-speaking countries such as the United States, Canada, the United Kingdom, and Australia. These sites tend to have a somewhat higher level of accessibility. However, universities in some non-English-speaking countries have significantly less accessible web pages. It is not clear from the present research what factors may cause this discrepancy in web site accessibility. It seems that legislation may have a limited positive effect on accessibility. For example, web sites in the United Kingdom, where university web site accessibility is regulated, have some of the highest levels of accessibility." (Kane, 154)

Dodd identifies that other usability concerns crop up as a result of the very standard for web pages (HTML, based off Dexter Model). "As a simple example, the ordering of content in a list may need to vary between visual and audio design spaces: in text-to-speech environments, the most important menu options may need to be listed first, but in a visual environment the menu may be required to be grouped more logically." (Dodd, page 8) This is a standard that is so long in evolution and entrenchment is so high that it is unlikely to change. " HTML version 5 [8] is now out for discussion and, “intends to replace HTML 4, XHTML 1.x, and DOM2 HTML specifications”" (Dodd from W3C, pg 10) Note that attributes and events can be used to provide synchronization (video.timeupdate event), but these require manual checking of the time every millisecond and is "forced synchronization".

Even with standard adherence, accessibility is not guaranteed. Screenreaders evaluate DOM at load-time and can not reevaluate when changed at run-time through on-page scripts. "Screen reading application Jaws for Windows [5] and VoiceOver [7], both fail to handle the dynamic page updates correctly on what is otherwise a trivial web page." (Dodd, pg 9)

CONCLUSION Too long an standardization time means continual adjusting or standards and implementations there of (ARIA, Firefox kept tweaking as standards) or non-implementation at all (IE held off on implementation). Too short can lead to short-sightedness, inflexibility and necessity to "go back to the drawing board" (WCAG 1 was narrow focused and tech specific. Implemented in 1999, version 2.0 was being drafted 2 years after already). Why would a company invest lots of resources in implementing WCAG 1 compliance a year after if WCAG 2 is already in the works (similar to IE not adopting ARIA, universities not conforming to WCAG "recommendations").