Testing for Accessibility

California Web Accessibility Conference

September 21-22, 2006

Jim Thatcher

jim@jimthatcher.com

Accessibility Consulting

Austin, Texas

---

Questions - comments

• Please don't hesitate to

  • ask questions at any time

  • make comments at any time

  • when I'm unclear, stop me

---

Outline

• Validating – Testing – Standards

• Testability and tools

  • Contrast (Contrast Analyzer)

  • Quality Link Text (Testing tools Report)

  • Text equivalents for non-text content (Toolbars and Favelets)

  • Forms

There is a spectrum of possibilities from validating to usability testing

---

Validating – Testing

• In April Course was called "Validating to 508"

• What's wrong with that title?

• You can "validate code to a standard" like to HTML 4.01, XHMTL 1.0 ()

• You cannot validate – in that sense – for accessibility

• There are "standards and guidelines" to which you may want to "validate"

---

Standards and Guidelines

• WCAG 1.0 Web Content Accessibility Guidelines Version 1.0 (16 P1, 30 P2, 19 P3)

• WCAG 2.0 (23 L1, 18 L2, 31 L3)

• Section 508 (16 provisions)

• Section 508 To be revised. Meeting today!

---

Testable? Automatic tests?

• Testable by a computer vs. requiring judgment

• claim compliant/accessible because no errors reported by testing tool

• Testing all for Section 508 Standards discussed

• Lets look at testability of some of the guidelines and some tools as we go

http://www.cyberseniors.org/artman/publish/index.shtml – example of compliancy claim.

---

Contrast – WCAG 1.0 & 2.0

• WCAG 1.0  Ensure that foreground and background color combinations provide sufficient contrast when viewed by someone having color deficits or when viewed on a black and white screen. [P2 for images, P3 for text]. (emphasis added)

vs.

• WCAG 2.0 1.4.1 Text or diagrams, and their background, have a of at least 5:1. (L2)

• WCAG 2.0 1.4.3 Text or diagrams, and their background, have a of at least 10:1. (L3)

WCAG 1.0 required "sufficient contrast" How could that possibly be tested – by computer or otherwise. It could not.

WCAG 2.0 is very specific – assuming you know what

Go to desktop and use color analyzer. Colour Contrast Analyzer (Google)

---

Tools: Contrast – Analyzer

• Contrast analyzer from WAT-C (Web Accessibility Tools Consortium)

• Check out the tool on or

• Samples:

WCAG 1.0 required "sufficient contrast" How could that possibly be tested – by computer or otherwise. It could not.

WCAG 2.0 is very specific – assuming you know what

Go to desktop and use color analyzer. Colour Contrast Analyzer (Google)

---

Automatic Contrast test?

• Could this be tested automatically?

  • Wouldn't be easy

  • But certainly feasible for text

  • Probably not for images

• Not done today (WCAG 2.0 too new and not final)

• Didn't look for this in the tools study,

WCAG 1.0 required "sufficient contrast" How could that possibly be tested – by computer or otherwise. It could not.

WCAG 2.0 is very specific – assuming you know what

Go to desktop and use color analyzer. Colour Contrast Analyzer (Google)

---

Link text - WCAG 1.0

• WCAG 1.0 13.1: Clearly identify the target of each link. (P2)

• Interpreted to mean good link text or title attribute on anchor element. See list of hotels on priceline.com

• It is testable – evaluators would usually agree on "good link text." Tools – lists of links.

• (Not 508 requirement)

---

Link text – automatic tests

• Raise errors or warnings for Common phrases, "more," "click here," "more …," "details" etc. Editable with several tools.

• Inadequate phrase like "click here" but with title attribute – Should pass

• Same phrase different href – should fail without title.

• Extreme case - Empty link text or spaces for link text

---

Specify unacceptable link text

Lift Machine from UsableNet

---

Testing tools report

• Looking at 6 tools and how they behaved with respect to 40 issues

• Like special link text ..

---

Desktop Commercial Tools

• Lift ($999)

• Ramp ($69-$1499)

• InFocus ($1795)

  • www.ssbtechnologies.com

• Bobby ($299)

---

Enterprise Tools (expensive)

• WebKing by Parasoft

• WebXM by Watchfire

---

Testing tools and link text

Details at

---

Link text – WCAG 2.0

• 2.4.4 Each link is programmatically associated with text from which its purpose can be determined. (L2) (emphasis mine)

  • Not testable – Key term not defined

• 2.4.8 The purpose of each link can be from the link. (L3)

  • Old interpretation the we just discussed.

  • Note Level 3

---

Discussion OF 2.4.4 (9/6/2006)

• On the WCAG Working Group mailing list

• … for HTML …the only acceptable context that can be added to the link text is

  • title attribute

  • enclosing sentence

  • enclosing paragraph

  • enclosing list item

  • enclosing table cell

  • header table cell

• When this gets resolved 2.4.4 will change (I hope) SO 2.4.4 will be testable

---

Text Equivalents – Standards

• 1194.22(a) A text equivalent for every non-text element shall be provided (e.g., via "alt", "longdesc", or in element content).

• Testability?

---

Text Equivalents – Standards

• WCAG 1.0 1.1 Provide a text equivalent for every non-text element (e.g., via "alt", "longdesc", or in element content). … This includes: images, graphical representations of text (including symbols), image map regions, animations (e.g., animated GIFs), applets and programmatic objects, ascii art, frames, scripts, images used as list bullets, spacers, graphical buttons, sounds (played with or without user interaction), stand-alone audio files, audio tracks of video, and video. (P1)

• We will look at WCAG 2.0 – but how do testing tools do

---

Text Equivalents – tools

• In the test of 6 tools, all detected that the sample file was missing alt text. Not a surprise

• Two of 6 called out an error with ASCII art as alt-text (Ramp and WebKing)

• The one thing testing tools should get consistently right is missing alt-text!

---

Choosing WCAG 1.1 - Bobby

Bobby from Parasoft

---

Choosing WCAG 1.1 InFocus

InFocus from The Bart Group (SSB)

---

Results for 1.1 on

---

Quality of alt-text

• So that is not very encouraging for automatic testing.

• What about the quality of alt-text

• How do we decide if the alt-text is good or adequate – links, pictures, cool scenes, captchas, image links …

• WCAG 2.0 again helps

---

Text Equivalents – WCAG2.0

• 1.1.1 For all , one of the following is true:

• If non-text content presents information or responds to user input, serve the same purpose and present the same information as the non-text content. If text alternatives cannot serve the same purpose, then text alternatives at least identify the purpose of the non-text content.

• …

---

Text Equivalents – WCAG2.0

• If non-text content is ; or content; a ; or primarily intended to create a ; then text alternatives at least identify the non-text content with a descriptive text .

• … more

---

Text Equivalents - WCAG 2.0

• If the purpose of non-text content is to confirm that content is being operated by a person rather than a computer, different forms are provided to accommodate multiple disabilities.

• More …

---

Text Equivalents – WCAG2.0

• If non-text content is pure decoration, or used only for visual formatting, or if it is not presented to users, it is implemented such that it can be ignored by assistive technology.

• That's it!!

---

Text Equivalents for what

• Look at images - what are these "images"

• Does every <img>, <area>, and input with type="image" have an alt attribute?

  • If "no" then page fails – tools can fail pages

  • If "yes" then not sure – need human evaluation

---

Toolbars - testing alt-text

• Toolbars – a great addition for accessibility

• Web accessibility toolbar from Vision Australia

• The Section 508 toolbar from RampWEB (IE or Firefox)

---

Toolbars 2 - testing alt-text

• The NCAM QA Favelet

• Web Developer Extension for Firefox

---

Alt-text

• When alt-text missing fails all guidelines and the HTML specification, but …

• Degrees of failure – what happens when alt is missing?

  • Active images – alt gives function of image

  • Information bearing images – alt gives information

  • Pure decoration or formatting images – alt="" (empty)

• Jim's Favelets for rapid human review

• Listen to !

---

Form Labeling – 508

• 1194.22 (n) When electronic forms are designed to be completed on-line, the form shall allow people using assistive technology to access the information, field elements, and functionality required for completion and submission of the form, including all directions and cues.

• How do we test? Does WCAG 1.0 help?

---

Form Labeling – WCAG 1.0

• 10.2 Until user agents support explicit associations between labels and form controls, for all form controls with implicitly associated labels, ensure that the label is properly positioned. (P2) (Emphasis mine)

• 12.4 Associate labels explicitly with their controls. (P2)

---

Testing Form Labeling

• Very important for screen reader users.

• Pretty simple and reliable machine testing:

  • Presence of label element or

  • Presence of title attribute on the input elements.

• And how do the testing tools do?

---

Form options with Ramp from DeQueue

---

Choosing Forms only Lift Machine from UsableNet

---

Lift – results with Forms

• travel.priceline.com/vacationpackages

• amazon.com/gp/yourstore/home

---

Ramp – results with Forms

---

Forms – with Jim's Favelets

• Try these

• • 88 form controls with 0 errors

• • 8 form controls with 8 errors

• Jim's Favelets:

---

Forms in WCAG 2.0

• 1.3.1 Information and relationships conveyed through can be , and notification of changes to these is available to , including assistive technologies. (L1)

• Techniques include …

---

Some techniques for SC 1.3.1

• Using label elements to associate text labels with form controls

• Using the title attribute to identify form controls when the label element cannot be used

• Providing a label for groups of radio buttons or checkboxes using the fieldset and legend elements

---

Free Web-based tools (single page )

• CynthiaSays (HiSoftware)

• WebExact (Watchfire)

• Site Valet (Standards –Shmandards)

• Hermish

Cynthia and site Valet use target and jt.com

---

Errors found by free tools

---

"Evaluation" of Free Tools

• Gez Lemon,

• Article

• My Response –

• Gez' test page – an excellent resource

---

More General Free Tools

• The Wave

  • wave.webaim.org

• A-Prompt

  • aprompt.snow.utoronto.ca

• UIUC Functional Accessibility Evaluator

  • A new and different entry

  • appserv.rehab.uiuc.edu/fae

---

Assistive Technology

• Testing by people with disabilities

• Testing with assistive technologies

  • HPR (IBM)

    • www.ibm.com/able/hpr.html

  • JAWS (Freedom Scientific)

  • Window-Eyes (GW Micro)

    • www.gwmicro.com

---

IBM Accessibility Testing
(Matt King et al)

• Monitoring checkable errors

• Human Review Component

---

Human Review for IBM

• Needed time per page around 3 minutes

• The toolbars don't do what testing tools do.

• Need value of software testing tools (no alt attribute)

• Plus ease of human evaluation

• My evaluation techniques

---

"Importance"

• Impact on people with disabilities

• Importance of the page

• Both are important – when talking about testing you must consider return on investment

• STEP (Simple Tool for Error Prioritization)

---

Six Commercial Tools,
A Comparison

• Chapter 13 – Reulatory Complaince …

  • See jimthatcher.com/testing.htm.

• Forty "issues"

  • jimthatcher.com/step/new

• Based on 70 files

Chapter 13 of the new book, Web Accessibility, Web standards and regulatory compliance.

SO you can read CH 13 on my web site.

---

Testing tools and link text

Details at

---

Thanks!

• Thanks for being here

• I can be reached at

• You can contact Knowbility

  • knowbility@knowbility.org