1 of 27

What do we mean when we talk about “scoping”?

2 of 27

Agenda

Problem statement
Examples of areas of exploration
Some attempts at terminology
Possible future discussions

3 of 27

Problem statement

In WCAG 2.x we used a URI ‘page’ to determine conformance. With changes in technology as well as a pivot to a more user-centered approach, we need to rethink the scope of conformance.

4 of 27

What is the smallest unit we test against?

Easy answer: the User Interface Component�“a part of the content that is perceived by users as a single control for a distinct function”
User perception and technology do not always align.
What about subcomponents?
What about operational affordances?
What about ‘composite’ components like ribbons/menu bars?

What about non-functional components? Text, graphs, sound, touch

The recent work with Focus Appearance raised some interesting discussions in this space.�- perception is not the same for everyone

- user perception and the mechanics ‘behind the scenes’ don’t always align

Lesson: we get value from a measurement and definition, even where it’s subjective

For FA, we had to examine the question of subcomponents.

It gets more complex thinking of things that are built out of components, but may themselves be perceived as a ‘thing’.

Some text, like a label, can be considered part of a component. Some seems like it’s more page or process-relevant – which we’ll touch on later..�Non-functional: Plain language

Another consideration in this space is that there are a lot of design and pattern libraries (APG) in this space.

Ultimately a key question is: what moves us towards repeatable, measurable, and consistent results?

5 of 27

What is the component?

Is a date input with no calendar a component?

Is the calendar part of an input, one input, or more than one?

6 of 27

Are 2.x criteria objective or subjective?

7 of 27

Objective test

A test that leads to consistent results with high* reliability

Measures on existence of something, normally against a constant

Numerical value (color contrast)
Presence of attribute/value (language attribute and value)

Repeatable, measurable, prescribed (hopefully easily)
2.x criteria aspire to be measurable through objective tests

8 of 27

Are 2.x criteria objective or subjective?

Looked at the first 2 dozen�- “objective only”, “both”, “subjective only”

Broke out wording into parts
ALL were “both”
Extended to first 3 dozen, across 5-point continuum

�Quick exercise in assessing WCAG 2.x for testing objectivity

9 of 27

1.4.2 Audio control: fully objective?

If any audio on a Web page plays automatically for more than 3 seconds, either a mechanism is available to pause or stop the audio, or a mechanism is available to control audio volume independently from the overall system volume level.

10 of 27

1.4.9 Images of Text (No exceptions)

Images of text are only used for pure decoration or where a particular presentation of text is essential to the information being conveyed.

“This does not include text that is part of a picture that contains significant other visual content.”

11 of 27

Exploration: Non-text content

All non-text content that is presented to the user has a text alternative that serves the equivalent purpose, except for the situations listed below.

Objective: “has a text alternative”
Subjective: “that serves the equivalent purpose”

Each of the 6 exception situations has its own considerations.

12 of 27

Objective tests in Non-text content

Is there non-text content?

Is there an image?

Is there something with an implicit image role? <img>
Is there something with an explicit image role? `role=“img”`

Does the image have an attribute for a text alternative?

Is there an `alt` attribute?
Is there an `aria-label` attribute?

Is the code good?

Does the attribute have a value?
Is the attribute valid?
Is the parsing correct?

13 of 27

What is the equivalent purpose?

“A picture paints a thousand words”

“dog”

“young dog”

“happy dog”

”alert dog with tongue hanging out”

“golden retriever”

“Emma’s first pet portrait”

14 of 27

What if we assessed for other criteria?

Accuracy

Alt=“cat”; singular/plural “dogs”

Sufficiency

- Minimum of X non-article words

Construction

Subject/object
Noun/verb

Context

- User process

Existence of image caption
Page content

15 of 27

Context makes a difference

16 of 27

A few ways of getting to more objective

A criteria only for images

“Images presented to the user have a brief, accurate text alternative”
Brief:2-5 non-article words
Accurate: correct in all details provided

A criteria for important images

“Important images are programmatically indicated as important and have an additional text alternative that serves the equivalent purpose”
Important Images: images which if removed cause the remaining content to be insufficient for the majority of users

Retain a Decoration, Formatting, Invisible exception

Can we gain insight in ways to parse existing requirements based on examinations of the techniques, tests and Understanding document?

“short text alternative”

“long text alternative”

There seems to me potential ways we could give guidance on image importance in many situations. “Images that are important to content meaning”

–

We already have some similar qualitative wording on the relevant importance of non-text content in Non-text-contrast “Visual information required to identify”’ “Parts of graphics required to understand the content’�

Is there any technical advantage to having an actual “decorative” attribute value? An empty string may mean a lot of things.

Is there an advantage to have more than one image descriptive value? Short name. long description. Say there was only a requirement for a short description for all but unimportant images

Decorative: “If non-text content is pure decoration, is used only for visual formatting, or is not presented to users, then it is implemented in a way that it can be ignored by assistive technology.”

Pure decoration: “serving only an aesthetic purpose, providing no information, and having no functionality”

This all needs RESEARCH. We could apply the same level of scrutiny to all the other aspects of 1.1.1 to look for other avenues of opportunity

17 of 27

Label In Name

“For user interface components with labels that include text or images of text, the name contains the text that is presented visually.”

Note: “A best practice is to have the text of the label at the start of the name.”

18 of 27

Objective Tests For Label In Name

Is it a user interface component?
Is there a label?

Does the label contain text?
Does the label contain images of text?

If there criterion is relevant:

Does the accessible name contain the visible text?
Does the accessible name contain the visible text but with additional words interspersed with it?
Does the accessible name contain the visible text’s words but in a different order?

Is the code good?

Is the name coded in an accessibility-supported way?
Is parsing correct?

19 of 27

Subjective Tests For Label In Name

The UIC has a label that includes images of text

It only has one image of text—does that mean it doesn’t have to conform?

20 of 27

Strategies to improve reliability

Scrutinize SC wording, Understanding, Techniques, Definitions + outside
Subjective criteria can be made more reliable by making tests and outcomes more granular
Conditions are subjective, but if agreed on, move tests ‘left’
Confine to specific technologies; components
Decision trees
Conditions can also be in definitions
Lists/examples can clarify a condition, but where to put them and what if they’re not exhaustive?

21 of 27

Terminology - scopes

Current term (Editor’s draft)	Concept	Suggested term	Notes/Other term options
Items	Smallest unit of reporting measurement for testing	Component	Item, unit. Or maybe a list: UIC, image…
Views	All content visually and programmatically available without a substantive change	Views	Page at URI or other, ‘container’
User processes	The things that affect a user's ability to do a thing	User processes	Allow us to assess content by user task, activity
Aggregate	Larger scope as defined by a collection of smaller scopes (views, user processes, etc.)	Aggregate	Whole site/app/thing

22 of 27

Terminology - test types

Current term (Editor’s draft)	Concept	Suggested term	Notes/Other term options
Unconditional	A test that leads to a consistent binary* outcome A test that leads to consistent results with high** reliability	Objective	It is always binary? Is it acceptable to be inconclusive? * Is it 100%? Can the measure be 100% but not the outcome?.
Conditional	A test that requires an educated judgment call on whether a condition has been met, and so has very good but not 100% reliability	Qualitative	Options: qualified, judged, fixed subjective
Procedural	[to come from the Protocol-subgroup folks]	[Evaluating procedures]	[to come from the Protocol-subgroup folks]
Conventional	A test that requires a criteria (language, design decision, etc) be defined by the tester before proceeding.	Test case	Context-dependent, site/application-specific

23 of 27

Exploration: Multiple Ways

"More than one way is available to locate a Web page within a set of Web pages except where the Web Page is the result of, or a step in, a process."

Requirements from the understanding doc

"The intent of this Success Criterion is to make it possible for users to locate content in a manner that best meets their needs. Users may find one technique easier or more comprehensible to use than another."

Note: This SC does not have any specific implementation requirement.

24 of 27

Multiple Ways: Objective or Subjective?

No objective aspects without context of the site/web app, as "More than one way is available" doesn't define the ways. Understanding documentation offers examples, but not a definitive list (a good thing!).

Examples of Subjective bits:

How difficult does it seem to get to each page?
Does one way have poor discoverability compared with another?
What about more than two ways?
"…except where the Web Page is the result of, or a step in, a process."

25 of 27

Multiple Ways: Test case

Multiple ways inherently requires context, so: Test case

Identify the ways the site offers to navigate to pages
Identify the individual pages and pages that make the starting point of a process
Cross-reference the two lists to ensure that you can navigate to all pages identified in step 2 exist in two or more ways from step 1�Note: Now an objective piece!

26 of 27

Multiple Ways: Example test cases

Both examples assume exhaustive, consistent nav leading to all pages

If a site has a site map, it includes all individual pages and pages that make the starting point of a process.
If the site has a search feature, you can find all individual pages and pages that make the starting point of a process by searching for them.

27 of 27

Other lines of query

Is there value in how many times something is done right?
Is “page” still viable for conformance?
Reporting isn’t covered in 2.x. What are the ramifications of incorporating in 3.0?
What would happen if 3.0 was scoped to HTML versus technology agnostic?
Can requirements be written so they don’t overlap?