W3C Accessibility Guidelines (WCAG) 3.0

Abstract

W3C Accessibility Guidelines (WCAG) 3.0 will provide a wide range of recommendations for making web content more accessible to users with disabilities. Following these guidelines will address many of the needs of users with blindness, low vision and other vision impairments; deafness and hearing loss; limited movement and dexterity; speech disabilities; sensory disorders; cognitive and learning disabilities; and combinations of these. These guidelines address accessibility of web content on desktops, laptops, tablets, mobile devices, wearable devices, and other web of things devices. The guidelines apply to various types of web content including static, dynamic, interactive, and streaming content; visual and auditory media; virtual and augmented reality; and alternative access presentation and control. These guidelines also address related web tools such as user agents (browsers and assistive technologies), content management systems, authoring tools, and testing tools.

Each guideline in this standard provides information on accessibility practices that address documented user needs of people with disabilities. Guidelines are supported by multiple outcomes to determine whether the need has been met. Guidelines are also supported by technology-specific methods to meet each outcome.

This specification is expected to be updated regularly to keep pace with changing technology by updating and adding methods, outcomes, and guidelines to address new needs as technologies evolve. For entities that make formal claims of conformance to these guidelines, several levels of conformance are available to address the diverse nature of digital content and the type of testing that is performed.

See WCAG 3 Introduction for an introduction and links to WCAG technical and educational material.

This section (with its subsections) provides advice only and does not specify guidelines, meaning it is informative or non-normative.

Plain language summary of Introduction

W3C Accessibility Guidelines (WCAG) 3.0 shows ways to make web content and apps usable by people with disabilities. WCAG 3 is a newer standard than the Web Content Accessibility Guidelines (WCAG) 2.
WCAG 3 doesn’t replace WCAG 2. WCAG 2 is used around the world and will still be required by different countries for a long time to come.
Meeting WCAG 2 at AA level means you will be close to meeting WCAG 3.0, but there may be differences.
We have developed labels for sections to tell you how confident we are that the content will not change. This lets you tell whether we are trying out an idea or whether we have put a lot of work into the topic and we don’t expect it to change. There are 5 levels:
- Placeholder - we are thinking about the topic and plan to work on it in the next few months.
- Exploratory - we have started work on this topic but still have many questions.
- Developing - we have general agreement, but not all the details are worked out.
- Refining - we think we have a good idea and want public feedback about it. We are ready to have a few people try it out and see how well it works.
- Mature - we think it is ready to publish and we want to know if we missed any rare examples.

End of summary for Introduction

What’s new in this version of WCAG 3?

This draft includes an updated list of the potential outcomes which we are exploring. The list of outcomes is longer than a listing of Success Criteria in WCAG 2.2 because the intent at this stage is to be as inclusive as possible of potential outcomes. The final set of outcomes in WCAG 3 will be different than what is in this draft. Outcomes will be added, combined, and removed. We also expect changes to the text of the Outcomes. Only some of the Outcomes will be required at the base level of conformance.

The purpose of publishing this initial list is to:

better understand the scope of needs,
request assistance in identifying gaps, and
request assistance locating and conducting research to validate or invalidate the outcomes listed.

We did not make changes to conformance related sections and we did not publish an updated WCAG 3 Explainer.

This specification presents a new model and guidelines to make web content and applications accessible to people with disabilities. The W3C Accessibility Guidelines (WCAG) 3.0 support a wide set of user needs, use new approaches to testing, and allow frequent maintenance of guidelines and related content to keep pace with accelerating technology change. WCAG 3 supports this evolution by focusing on the functional needs of users. These needs are then supported by outcomes and technology-specific methods to meet those needs.

WCAG 3 is a successor to Web Content Accessibility Guidelines 2.2 [WCAG22] and previous versions, but does not deprecate WCAG 2. It will also incorporate content from and partially extend User Agent Accessibility Guidelines 2.0 [UAAG20] and Authoring Tool Accessibility Guidelines 2.0 [ATAG20]. These earlier versions provided a flexible model that kept them relevant for over 15 years. However, changing technology and changing needs of people with disabilities have led to the need for a new model to address content accessibility more comprehensively and flexibly.

There are many differences between WCAG 2 and WCAG 3. The WCAG 3 guidelines address accessibility of web content on desktops, laptops, tablets, mobile devices, wearable devices, and other web of things devices. The guidelines apply to various types of web content including static, dynamic, interactive, and streaming content; visual and auditory media; virtual and augmented reality; and alternative access presentation and control. These guidelines also address related web tools such as user agents (browsers and assistive technologies), content management systems, authoring tools, and testing tools.

Content that conforms to WCAG 2.2 A and AA is expected to meet most of the minimum conformance level of this new standard but, since WCAG 3 includes additional tests and different scoring mechanics, additional work will be needed to reach full conformance. Since the new standard will use a different conformance model, the Accessibility Guidelines Working Group expects that some organizations may wish to continue using WCAG 2, while others may wish to migrate to the new standard. For those that wish to migrate to the new standard, the Working Group will provide transition support materials, which may use mapping and other approaches to facilitate migration.

As part of the WCAG 3 drafting process each normative section of this document is given a status. This status is used to indicate how far along in the development this section is, how ready it is for experimental adoption, and what kind of feedback the Accessibility Guidelines Working Group is looking for.

Placeholder: This content is temporary. It showcases the type of content or section to expect here. All of this is expected to be replaced. No feedback is needed on placeholder content.
Exploratory: The working group is exploring what direction to take with this section. This content is not refined, details and definitions may be missing. Feedback should be about the proposed direction.
Developing: There is rough agreement on what is needed for this section, although not all high-level concerns have been settled. Details have been filled, but are yet to be worked out. Feedback should be focused on ensuring the sections are usable and reasonable in a broad sense.
Refining: The working group has reach consensus on this section. It is ready for broad public review and experimental adoption. Feedback should be focused on the feasibility and implementability.
Mature: Content is believed by the working group to be ready for recommendation. Feedback should be focused on edge case scenarios the working group may not have anticipated.

This section (with its subsections) provides requirements which must be followed to conform to the specification, meaning it is normative.

Plain language summary of Guidelines

The following guidelines are being considered for WCAG 3. They are currently a list of topics which we expect to explore more thoroughly in future drafts. The list includes current WCAG 2 guidance and additional requirements. The list will change in future drafts.

End of summary for Guidelines

Editor's note

The individuals and organizations that use WCAG vary widely and include web designers and developers, policy makers, purchasing agents, teachers, and students. To meet the varying needs of this audience, several layers of guidance will be provided including functional categories of disabilities, general guidelines, outcomes that can be tested, a rich collection of methods, resource links, and code samples.

The following guidelines are an initial list of potential outcomes that the working group will be exploring. The goal is to guide the next phase of work. They should be considered drafts and should not be considered as final content of WCAG 3.

Ordinarily, exploratory content includes editor's notes listing concerns and questions for each item. Because this Guidelines section is very early in the process of working on WCAG 3.0, this editor's note covers the content of the entire section. Please consider all items in the list as exploratory at this point. It is a list of all possible topics for consideration. Not all items listed will become requirements.

The outcomes listed below came from analysis of user needs that the Working Group has been studying, examining, and researching. They have not been refined and do not include essential exceptions or methods. Some outcomes may be best addressed by authoring tools or at the platform level. Many outcomes need additional work to better define the scope and to ensure they apply correctly to multiple languages, cultures, and writing systems. We will address these questions as we further explore each outcome.

Please consider the following questions when reviewing this list of outcomes:

What outcomes are needed to make web content accessible and are missing from this list?
What research supports or refutes these outcomes?
Are outcomes listed here out of the scope of accessibility guidance and why?

To provide feedback, please file a GitHub issue or send email to [email protected] (comment archive).

Additional Research

One goal of publishing this list is to identify gaps in current research and request assistance filling those gaps.

Editor's notes indicate the outcomes within this list where the working group has not found enough research to fully validate the guidance and create methods to support it or additional work is needed to evaluate existing research. If you know of existing research or if you are interested in conducting research in this area, please file a GitHub issue or send email to [email protected] (comment archive).

End of note

Audio shifting designed to create a perception of motion is avoided; or can be paused or prevented.

Editor's note

Needs additional research

End of note

Flashing or strobing beyond thresholds defined by safety standards are avoided; or can be paused or prevented.

Visual motion and pseudo-motion after a specified time is avoided; or can be paused or prevented.

Editor's note

Needs additional research

End of note

Automated input from user agents, 3rd party tools, or copy-and-paste is supported.

Error notifications are programmatically associated with the error source.

Errors are visually identifiable without relying on only text, only color, or only symbols.

Error notifications are provided when an error occurs that describe the error and either provide instructions to fix the error or state that the system is at fault.

Input constraints or conditions (required line length, date format, password format, etc) are programmatically and visually indicated.

Inputs have visible labels that identify the purpose of the input.

Error notifications persist until the user dismisses them or the error is resolved.

Error notifications are visually collocated with the source of the error within the viewport, or provide a link to the source of the error which, when activated, moves the viewport to the error.

Editor's note

Needs additional research

End of note

Enough time is provided to read and use content.

Editor's note

Needs additional research

End of note

The interface indicates when user input or action is required to proceed.

Tasks can be completed without navigating misinformation or redirections.

Editor's note

Needs additional research

End of note

The interface supports stepping backwards in a process and returning to the current point without data loss.

Information needed to complete a multi-step process is provided at the start of the process, including:

estimated time or number of steps it might take,
details of any resources needed to perform the task, and
overview of the process and next step.

Tasks, including login/authentication, can be completed without puzzles, calculations, or other cognitive tests (essential exceptions would apply).

Tasks can be completed without memorizing and recalling information from previous stages of the process.

Editor's note

Needs additional research

End of note

Tasks can be completed without reading or understanding unnecessary content.

Tasks can be completed without entering unnecessary information.

Preselections are visible during task completion.

Data entry and other task completion processes allow saving and resuming from the current step in the task.

The steps and directions needed to complete a process are visually and programmatically indicated.

For tasks where labels and instructions cannot provide sufficient instructions for completion, detailed documentation on task completion is available.

Editor's note

Needs additional research

End of note

Auto generated text descriptions are editable by content creator.

Editor's note

Needs additional research

End of note

Equivalent audio alternatives are available in the preferred language.

Equivalent visual alternatives are available as synchronized audio in the media.

Equivalent audio alternatives are available as synchronized captions in the media.

Equivalent text alternatives are available for complex images.

Image alternatives include context.

Editor's note

Needs additional research

End of note

Equivalent descriptive text alternatives are available for decorative images.

Needs additional research

Equivalent audio and visual alternatives to audio and video alternatives are available in descriptive transcripts.

Media that has the desired media alternatives (captions, audio descriptions, and descriptive transcripts) can be found.

Editor's note

Needs additional research

End of note

Auto generated text alternatives are identified.

Editor's note

Needs additional research

End of note

Equivalent text alternatives are available for images that convey content.

The role and importance of images are programmatically indicated.

Editor's note

Needs additional research

End of note

The type of image (photo, icon, etc.) is indicated.

Editor's note

Needs additional research

End of note

Equivalent text alternatives are available for images of text.

Equivalent text alternatives are available for non-text, non-image content that conveys context or meaning.

Media alternatives explain nonverbal cues, such as tone of voice, facial expressions, body gestures, or music with emotional meaning.

Editor's note

Needs additional research

End of note

Figure captions persist or can be made to persist.

Editor's note

Needs additional research

End of note

Information is not conveyed with color alone.

Information is not conveyed with visual depth alone.

Editor's note

Needs additional research

End of note

Information is not conveyed with sound alone.

Information is not conveyed with spatial audio alone.

Controls and inputs with the same functionality behave consistently.

Selecting an element with a ‘pointer’ sets the focus to that element.

Controls have visible labels that identify the purpose of the controls.

Controls follow established conventions.

Editor's note

Needs additional research

End of note

Controls and inputs with the same functionality have consistent labels.

The importance of controls is visually and programmatically indicated.

Editor's note

Needs additional research

End of note

Changes to control or input name, roles, values or states are visually and programmatically indicated.

Controls and interactions are not deceptively designed (invisible, incorrectly labeled, placement, etc.).

Editor's note

Needs additional research

End of note

Controls are visually distinct from static content and include visual cues on how to use them.

Components which trigger a 'change of context' are indicated, or the change of context can be reversed.

Interaction indicators meet a ‘minimum contrast ratio text’ and meet a minimum thickness.

Editor's note

Needs additional research

End of note

Interactive components are available to all navigation and input methods.

Accurate names, roles, values, and states are available for interactive components.

Visual information required to identify user interface components and states meet a ‘minimum contrast ratio test’, except for inactive components or where the appearance of the component is determined by the user agent and not modified by the author.

Editor's note

Needs additional research

End of note

Notification is provided when previously viewed content changes.

Controls that can alter the order of content convey their purpose prior to activation, and convey their impact on content order when activated.

The focus or point of regard is restored to its previous location after a temporary change of view.

The focus order does not include repetitive, hidden, or static elements.

All functionality can be used without needing to accurately position a pointer.

Controls that have similar function and behavior have a consistent visual design.

The number of input commands required to complete a task using the keyboard is similar to the number of input commands when using other input modalities.

Editor's note

Needs additional research

End of note

Keyboard interface interactions are consistent.

The focus does not move to a position outside the current viewport, unless a mechanism is available to return to the previous focus point.

Every function that can be operated by a pointer, can be operated by a ‘single pointer input’ or a sequence of single pointer inputs, without limitations on timing for input.

Application keyboard commands do not conflict with platform commands, and the user is informed of non-standard commands.

The keyboard focus is visually indicated.

All functionality can be performed through the keyboard interface only, except where the underlying function requires input that depends on the path of the user's movement and not just the endpoints.

If keyboard focus can be moved to an interactive component, then the keyboard focus can be moved away from that component.

Functionality which supports pointers can be used by any pointing device supported by the platform.

Pointer cancellation is consistent.

Users are able to determine where the pointer is located.

Click activation using a pointer device does not require applying a specific pressure.

Editor's note

Needs additional research

End of note

Use of a pointer does not require a particular speed of pointer movement or click activation.

The keyboard input mode is indicated.

All functionality can be done without needing to move their body, except for accessibility supported input devices.

All functionality can be done without needing to move the hardware device.

Any input modality available on a platform can be used concurrently.

The relationships between parts of the content is clearly indicated.

The starting point or home is visually and programmatically labeled.

The source of the interface and primary content is visually and programmatically indicated.

Editor's note

Needs additional research

End of note

The order of content and interactions remain consistent throughout a workflow.

Content orientation allows the user to read the language presented without changing head position.

Controls are visually and programmatically located in an expected location.

Editor's note

Needs additional research

End of note

The current location within the view, multi-step process, and product is visually and programmatically indicated.

Meaningful associations between distinct pieces of content are programmatically determinable.

Editor's note

Needs additional research

End of note

Sections are visually and programmatically distinguishable.

Editor's note

Needs additional research

End of note

Common components and layouts are used.

A user can focus on a content “area,” such as a modal or pop=up, then resume their view of all content using a limited number of steps.

Third party content (AI, Advertising, etc.) is visually and programmatically indicated.

Editor's note

Needs additional research

End of note

When interfaces dramatically change (due to redesign), a mechanism to learn the new interface or revert to the older design is available.

Editor's note

Needs additional research

End of note

Provides context that orients the user in a site or multi-step process.

Changes in content and updates notify users, regardless of the update speed.

Related information is grouped together within a visual and programmatic structure.

Provides context that orients the user in a site or multi-step process.

Interactive components retain their position unless a user changes the viewport or moves the component.

Sections of content have well structured, understandable visual and programmatic headings.

Content is organized into short sections of related content.

Editor's note

Needs additional research

End of note

The purpose of each section of the content is clearly indicated.

Use does not cause visual overstimulation.

Editor's note

Needs additional research

End of note

Whitespace separates chunks of content.

The product provides at least two ways of navigating and finding information (Search, Scan, Site Map, Menu Structure, Breadcrumbs, contextual links, etc).

Algorithms (including AI) used are not biased against people with disabilities.

Editor's note

Needs additional research

End of note

The interface clearly indicates when a user is entering an agreement or submitting data.

Editor's note

Needs additional research

End of note

Disability information is not disclosed to or used by third parties and algorithms (including AI).

Editor's note

Needs additional research

End of note

Task completion does not include exploitive behaviors.

Editor's note

Needs additional research

End of note

A mechanism is available to prevent fraudulent redirection or alert users they are exiting the site.

Editor's note

Needs additional research

End of note

Prompts to hide and remove sensitive information from observers are available.

Editor's note

Needs additional research

End of note

The expanded form or meaning of abbreviations and acronyms is available.

Alternative formats for ambiguous number formats are available.

All letters and diacritics needed to phonetically read words are available.

The language and tone used is respectful and appropriate to the topic or subject matter.

Editor's note

Needs additional research

End of note

When font size conveys meaning (such as headings), the text maintains its meaning and purpose when text is resized.

Content does not include double negatives to express a positive unless it is standard usage for that language or dialect.

Explanations for figurative and non-literal language [such as jokes, sarcasm, hyperbole, metaphors, similes, and idioms] are available.

The interface avoids overwhelming verbosity.

Editor's note

Needs additional research

End of note

Three or more items of related data are presented as bulleted or numbered lists.

The rendered text against its background meets a maximum ‘contrast ratio test’ for its text appearance and use.

Editor's note

Needs additional research

End of note

The rendered text against its background meets a minimum ‘contrast ratio test’ for its text appearance and use.

Editor's note

Needs additional research

End of note

Steps in a multi-step process are numbered.

Clear explanations of the risks and consequences of choices, including use, are stated.

Editor's note

Needs additional research

End of note

Meaning conveyed by text appearance is programmatically available.

Editor's note

Needs additional research

End of note

The voice used is easiest to understand in context.

Editor's note

Needs additional research

End of note

Each segment of text [such as sentence, paragraph, bullet] presents one concept.

Access to a plain-language summary, abstract, or executive summaries is available.

Text or visual alternatives are available for numerical concepts.

The rendered text meets a minimum font size and weight.

Editor's note

Needs additional research

End of note

The rendered text does not use a decorative or cursive font face.

Text content can be converted into speech.

Editor's note

Needs additional research

End of note

Content has a title or high-level description.

Each paragraph of text begins with a topic sentence stating the aim or purpose.

Definitions for uncommon or new words are available.

Editor's note

Needs additional research

End of note

Sentences are concise, without unnecessary filler words and phrases.

The verb tense used is easiest to understand in context.

Editor's note

Needs additional research

End of note

Help is labeled consistently and provided in a consistent visual and programmatic location.

Editor's note

Needs additional research

End of note

Contextual help is available.

Conversational support is available that allows both text and verbal communication.

Help understanding and using data visualizations is available.

Editor's note

Needs additional research

End of note

Feedback can be provided.

Help using new or changed interfaces is available.

Editor's note

Needs additional research

End of note

Adaptable/personalizable help is available.

Editor's note

Needs additional research

End of note

Instructions and help do not rely on sensory characteristics.

Visual illustrations, pictures, and images are available to help explain complex ideas, events, and processes.

Editor's note

Needs additional research

End of note

Accessible support during data entry, task completion and search is available.

Editor's note

Needs additional research

End of note

Text and background colors can be customized.

Complex information or instructions for complex processes are available in multiple presentation formats.

Editor's note

Needs additional research

End of note

Content can be controlled using assistive and adaptive technology.

Audio can be turned off, independent of the system audio, while allowing video to play.

The position and formatting of captions can be changed.

Large amounts of data can be broken into smaller chunks.

Patterns, designs or images placed behind text are avoided or can be removed.

The timing and positioning of notifications and other interruptions can be changed, suppressed or saved, except interruptions involving an emergency.

Warnings about content that may be emotionally disturbing are available and the content can be hidden.

Editor's note

Needs additional research

End of note

Haptic feedback can be reduced or turned off.

Editor's note

Needs additional research

End of note

The ability to look up terms within audio alternatives is available.

Editor's note

Needs additional research

End of note

Captions and audio descriptions can be turned on and off.

Media can be navigated by chapters.

Editor's note

Needs additional research

End of note

Printing respects user’s content presentation preferences.

Editor's note

Needs additional research

End of note

Content can be viewed in multiple size viewports, orientations, and zoom levels without loss of content, functionality and meaningful relationships and with scrolling only occurring in 1 direction.

Text appearance [font, size, etc] and layout [spacing, single column] can be customized by the user.

Advertising and other third-party content that obscures the primary content can be moved or removed before interacting with it.

Editor's note

Needs additional research

End of note

Content can be transformed to make its purpose clearer.

Editor's note

Needs additional research

End of note

Triggering content is indicated and the content and trigger warnings can be hidden.

Editor's note

Needs additional research

End of note

User settings are honored when using or reviewing content.

Assistive technologies can access content and interactions when mechanisms that convey alternative points of regard or focus (i.e. virtual cursor) are used.

This section (with its subsections) provides requirements which must be followed to conform to the specification, meaning it is normative.

Plain language summary of Conformance

You might want to make a claim that your content or product meets the WCAG 3 guidelines. If it does meet the guidelines, we call this “conformance.” To conform to WCAG 3, your test results must show that your project is accessible.

If you want to make a formal conformance claim, you must use the process described in this document. However, conformance claims are not required and your content can conform to WCAG 3, even if you don’t want to make a claim. You can still use this process to test your project’s accessibility.

There are two types of content in this document:

Normative: what you must do to meet the guidelines.
Non-normative: advice to help you meet the guidelines. This is also called informative.

There are a variety of ways to say what is required in WCAG 3. We are experimenting with different approaches. Once we have developed enough guidelines, we will test how well each works. Here are some of the ideas for WCAG 3 requirements:

Two ways of evaluating accessibility:
- Outcomes are statements that can be tested.
- Assertions and procedures are statements that you used an accessible way to create the content that is hard to test. Assertions are new and we are trying different rules for how they can work.
Outcomes have two (2) types of tests:
- Quantifiable tests: Tests where there is a high degree of consistency between test results from different testers. Examples include testing whether certain properties exist in the content or if they match a value specified by the requirement.
- Qualitative tests: Tests that rely on a qualitative evaluation based on existing criteria. Test results may vary between testers who understand the criteria. An example is evaluating the quality of a requirement such as alternative text.
Tests can be applied to four (4) different scopes:
- Item: A component or unit of content. Examples include a drop down menu, a media player, a phrase, or an image.
- View: All content visually and programmatically available without a user-initiated substantive change.
- User process: Series of user actions, and the distinct items and interactive views that support the actions, where each action is required to complete an activity.
- Product: The aggregate of all related items, user processes, and views.
Three different levels show how accessible the content is: bronze, silver, or gold level.
Issue severity allows different rules for accessibility problems. Rules for problems that block people from using the content are stronger than rules for minor problems.
Adjectival ratings allow testers to grade content using words (fail, pass, great) or numbers (1,2,3). This provides a scale that may better match what people with disabilities go through to use it.
Percentages are used instead of requiring that accessibility be perfect or you fail.
Pre-assessment conditions are tests that coders can use to see if they are ready to start accessibility testing. This is to help coders, not to make a new level of requirements.

We intend to address the following topics in the future, but for now, you can skip them.

Conforming alternative version - whether WCAG 3 will allow having inaccessible content if there is also an accessible version.
Only accessibility-supported ways of using technologies - this is a complicated topic about different assistive technologies, their features, and the languages they are available in.
Defining conformance scope - WCAG 2.x defines scope as a web page. WCAG 3 will have a broader scope to include the modern web and applications.
Conformance requirements - what someone has to do to meet WCAG 3.
Conformance claims – how to make a statement that you meet WCAG 3.

End of summary for Conformance

In addition to this section, the Guidelines and Conformance sections in WCAG 3 provide normative content and define requirements that impact conformance claims. Introductory material, appendices, sections marked as non-normative, diagrams, examples, and notes are informative (non-normative). Non-normative material provides advisory information to help interpret the guidelines but does not create requirements that impact a conformance claim.

The key words MAY, MUST, MUST NOT, NOT RECOMMENDED, RECOMMENDED, SHOULD, and SHOULD NOT in this document are to be interpreted as described in BCP 14 [RFC2119] [RFC8174] when, and only when, they appear in all capitals, as shown here.

WCAG 3 will include a new conformance model to address a wider range of user needs, test a wider range of technologies and support new approaches to testing. We are exploring several approaches to conformance. After studying the comments on the previous draft, these are the concepts that showed promise. We are giving an overview in this draft, but we continue to test the combination of the concepts.

There are several goals for this new conformance model:

Develop a model that encourages web sites to continue to improve accessibility (vs. stopping at the previous AA level);
Better reflect the lived experience of people with disabilities, who successfully use sites that have some content that does not meet WCAG 2.0 AA, or who encounter barriers with sites that meet WCAG 2.0 AA; and
Allow for bugs and oversight by content authors, provided the impact of them upon users with disabilities is not substantial.

The proposed approaches can fit together in a variety of ways. We will be testing these approaches and others for validity, reliability, sensitivity, adequacy, complexity and equity. We welcome suggestions on ways to improve these approaches to better meet these criteria and concerns about how they might affect accessibility. The working group plans to select from or even replace these options based on feedback, prototyping, and testing.

There are two main approaches to evaluating accessibility that are promising. There are also detailed ideas that support these approaches. The two main approaches are:

Outcomes with tests: Outcomes are verifiable statements that allow testers to reliably determine if the content being evaluated satisfies the user needs identified in the Guideline. Outcomes are addressed in Section 3.3.1 Outcomes. Tests are addressed in Section 3.3.2 Testing Outcomes.
Assertions and procedures: Assertions are attributable statements by a person or organization that they followed a procedure to improve accessibility. Assertions are addressed in Section 3.4.1 Assertions.

There are additional ideas that support these two approaches and can be used or combined in many different ways.

Conformance levels: Bronze, Silver, and Gold levels continue to receive positive feedback as an approach for overall conformance rating.
Issue severity: Outcomes may allow for the concept of varying severity. High severity issues are those which prevent users from completing user processes (tasks).
Adjectival ratings: Adjectival Ratings allow testers to grade a test, outcome, or guideline by an adjective rating (such as fail, pass, exemplary) or a numeric rating (such as 1-5) that potentially can be closer to the lived experience of a person with a disability.
Pre-assessment checks: Pre-Assessment checks are tests or criteria that implementers can use to determine if they are ready to assess conformance. The intent is to help organizations prepare for conformance testing, not to create a new level of conformance.

The details of these approaches change as we assemble them into a coherent whole. This draft gives a high level overview of these approaches so we can give an update and receive feedback on the individual approaches we are considering.

Editor's note

As we continue developing conformance, we seek input on the following:

Which option has the best chance of adoption and why?
How well do these approaches support regulatory needs?
How will these approaches be integrated into a conformance model (including levels or scores)?

Next steps include:

Further refine options,
Test the validity, reliability, sensitivity, adequacy, complexity and equity of the various models using these approaches, and
Write sample guidelines to test out each option.

End of note

Editor's note

As we continue developing outcomes and methods, we seek input on how well the approach to outcomes, assertions and tests defined here supports additional requirements not addressed in 2.2.

Next steps include:

Get feedback from designers, developers, and other communities on wording choice,
Finalize names and descriptions of scope and tests,
Develop detailed examples of methods and tests,
As we develop example outcomes and methods, further explore conditions and how multiple measurements might be used to meet an outcome.
Address all GitHub issues under test types and terminology milestone.

End of note

Outcomes are verifiable statements that allow testers to reliably determine if the content being evaluated satisfies the user needs identified in the Guideline. All outcomes and assertions that relate to a Guideline will be listed together to encourage adoption of higher levels of accessibility.

Each outcome is associated with at least one method. Methods are informative and kept in how to documents. Each method contains techniques for meeting the outcome, examples, resources, and sets of tests for evaluating the outcome. Methods can apply to a specific technology, such as HTML, or can be more generic where the advice applies no matter what technology, such as the methods supporting the Clear Language guideline.

Outcomes are written so that testers can determine the accessibility of technologies based solely on the outcome, even when methods do not yet exist for those technologies.

WCAG 3 includes two (2) types of tests which are evaluated:

Quantifiable tests: Tests where there is a high degree of consistency between test results from different testers. Examples include testing whether certain properties exist in the content or if they match a value specified by the requirement.
Qualitative tests: Tests that rely on a qualitative evaluation based on existing criteria. Test results may vary between testers who understand the criteria. One example is evaluating the quality of alternative text used to meet a requirement for alternative text.

Most tests have prescribed ways to meet the test. In some cases, the ways to meet the test will change based on a specific condition being met (example: the human language of the content).

Although content may satisfy all outcomes using quantifiable and qualitative tests, the content may not always be usable by people with a wide variety of disabilities. The assertions (see Section 3.4.1 Assertions) are designed to address this problem.

Quantifiable tests rely on measuring properties of the content based on nominal values. The test results are objectively verifiable, and avoid variation of test results between different testers. Values are quantifiable. They could be boolean (true/false), for example to check the presence of titles, text alternatives, and accessible names. Other values could include numerical thresholds; for example, to check color luminosity ratios.

Each method using quantifiable tests includes:

the values being tested; and
an algorithm to measure the properties of the content based on the values.

Qualitative tests rely on evaluating content based on a set of defined qualities and exceptions. The set of qualities and exceptions limit the scope of decisions, to minimize variation of test results arrived at by different testers. Still, some level of qualitative assessment is required, therefore the accuracy of the test results also depends on the knowledge and context of the testers to some degree.

Each method using qualitative tests includes:

the defined qualities being tested; and
guidance on evaluating how well the content meets the defined qualities.

Testing outcomes use items, views, user processes, and the product to define what is being tested.

Items are the smallest testable unit. They may be interactive components such as a drop down menu, a link, or a media player. They may also be units of content such as a word, a phrase, a label or error message, an icon, or an image.

Views include all content visually and programmatically available without a substantive change. Conceptually, views correspond to the definition of a web page as used in WCAG 2, but are not restricted to content meeting that definition. For example, a view could be considered a “screen” in a mobile app or a layer of web content – such as a modal.

User processes are a series of user actions, and the distinct interactive views and items that support the actions, where each action is required to complete an activity. A user process may include a subset of items in a view or a group of views.

Examples of a process include:

Logging into a web site and being recognized as an authenticated user;
Ordering an item, in which case the process includes the entire set of tasks from searching for the item, adding it to the shopping cart, paying for it, and receiving confirmation;
Submitting tax information, from start to end of the process; and
Interacting with other users in a virtual reality environment.

A process is comprised of one or more views or subsets of views. Only the part of the views that support the user process are included in a test of the user process.

The product is the combination of items, views, and user processes that collectively comprise the web site, set of web pages, web app, etc.

Some tests only apply in certain situations. Testing may occasionally require determining and referencing which specifications are being tested. Methods will note whether a test always applies or under what conditions a test applies. Both quantitative and qualitative tests can be conditional.

Editor's note

As we continue developing this content, we seek input on the following:

Can assertions be used to record accessibility work that is not required in the guidelines? This could include advance work on guidance not yet added to the guidelines.
What optional supporting documentation should organizations provide with an assertion?
Is there a need for WCAG 3 to require proof of an assertion, and if so, what documentation should be required as proof?
Should assertions be dated, expire, or be reviewed on a regular basis?
Can steps in a procedure duplicate tests in other parts of the guidelines? If so, how should those be handled?
Can assertions exist outside of conformance? For example, can they be used as an internal benchmark rather than for a claim of conformance?
Can assertions be used at the most basic level of conformance? If so, how?
How can small organizations use assertions without unrealistic burden?
As written, outcomes and assertions are at the same level. Would moving assertions to the test level be more effective?
The AGWG is considering whether and how assertions can be applied to the Bronze level.
The AGWG is considering what will qualify as a procedure in WCAG 3. A procedure may be limited to guidance:
- approved by AGWG and listed in WCAG,
- that references publicly published guidance, or
- that meets criteria specified in WCAG
or it may be any process that an organization uses to improve accessibility. Comments or criticism of these alternatives is welcome.

Next steps include:

Better define procedure,
Develop detailed examples of methods and test,
Test the accuracy, reliability, repeatability, etc. of this approach

End of note

An assertion is a formal claim of fact, attributed to a person or organization. In WCAG 3, an assertion is an attributable and documented statement of fact regarding procedures practiced in the development and maintenance of the content or product to improve accessibility.

Assertions may supplement methods in one or more outcomes. Assertions should only be used on outcomes and guidelines that allow assertions. Organizations can make an assertion that they followed a procedure to claim conformance. Results when testing assertions are true/false - the organization making the assertion either provided the required documentation or it did not.

Procedures used in assertions may be implemented at the organization level, during design and development, or during testing.

Examples of procedures that may be used during implementation might include:

Training,
FTE (Full Time Equivalent) assignments,
Skills testing,
Coordination and documentation of accessibility processes, or
Setting the priority for remediation.

Examples of procedures that may be used to evaluate accessibility might include:

Usability testing,
Heuristic evaluation, or
Assistive technology testing.

Assertions must be documented as part of the conformance claim process. The required information may also be made available through the web site.

Assertions might include the following information:

The statement being asserted,
The date of the assertion,
The date or date range the procedure was completed,
The scope of the assertion,
Contact information for the person or group making the assertion, and
The outcome(s) or guideline(s) supported by the assertion.

Editor's note

An alternative to specifying assertions at the outcome or guideline level might be to require the assertion apply to the scope of the conformance claim.

End of note

WCAG recommends maintaining additional information that an organization can use to improve or validate procedures and assertions. WCAG will not require organizations to provide supporting documentation to conform.

The quality of an assertion can be tested based on how well the assertion meets the documentation requirements for assertions (See Documenting Assertions). Conforming to WCAG does not require testing supporting documentation; however, organizations may decide to adopt additional documentation requirements based on the procedure being asserted.

WCAG 3 defines three levels of conformance: bronze, silver, and gold. While it is easy to replicate the WCAG 2 A, AA, AAA by renaming the levels, there is an opportunity to improve accessibility for people with disabilities by using a more advanced approach.

Bronze is the minimum conformance level. Content that does not meet the requirements of the bronze level does not conform to WCAG 3. To reach Bronze level, the scope claimed in the conformance statement must pass a subset of outcomes and assertions. The subset will require enough outcomes and assertions to improve equity across functional needs.

Silver level incentivizes organizations to go further to improve accessibility. One possibility that we are examining is that Silver level points can accumulate even prior to completing bronze but are not usable until Bronze is achieved. The goal is to encourage organizations to go beyond the minimum, especially where organizations want to be recognized for their efforts to go beyond minimum accessibility.

Gold level identifies measures we want to include for those organizations that do achieve Silver so that some can stand out as exemplary, cutting edge, and role models. There are a number of ideas that will be developed further once more of the conformance structure is solidified.

Editor's note

Severity rating could contribute towards scoring and prioritization.

As we continue developing this content, we seek input on the following:

Is every issue critical to someone, making this concept invalid?
How best to assign severity, particularly if testers have different ideas on what is critical?
How do we incorporate context/process/task? Is that part of scoping, or issue severity? Both are important to the end result.
What to do with non-critical issues?
If included, how will situations where severity depends on context be handled?
Can the matrix inform designation of functional categories? For example, the Text Alternative Available outcome.
How will issue severity fit into levels? For example:
- “Bronze” could be an absence of any critical or high issues;
- “Silver” could be an absence of any critical, high, or medium issues.
How to account for cumulative issues becoming critical?
Would another approach be more effective, for example assigning critical issues after testing is complete based on task or type of task rather than by test?

Next steps include:

Testing the assumption that some failures cause a greater impact to users than others or whether all guidelines and contexts are important to some individuals.
Explore whether the concept of issue severity can be applied consistently and effectively.

End of note

Outcomes may allow for the concept of varying severity. High severity issues are those which prevent users from completing user processes (tasks).

Tests could include critical issues. Each test could have a category of severity, so some tests will be flagged as causing a critical issue.

Adjectival Ratings allow test results to go beyond Pass or Fail to show progress towards a goal or exceeding a goal. Example of Possible adjectival ratings are:

Fail, Pass, Exceptional; or
Fail, Progress, Pass, Better, Exceptional.

Outcomes or guidelines could be evaluated using adjectival ratings on both directly quantifiable outcomes and qualitative measures that are asserted. Outcomes might be assigned an adjectival rating based on methods used to meet the outcome and issue severity. Guidelines might be assigned an adjectival rating based on the outcomes and assertions completed under the guideline.

Editor's note

We are exploring whether percentages could apply to Bronze but have not found a model to date where this works without adding complexity and time needed for testing.

As we continue developing this content, we seek input on the following:

How can percentages be used in a way that is equitable across disabilities?

End of note

In this approach, percentage of outcomes and assertions passed or percentage passed at a certain adjectival rating might be used to conform to Silver and Gold levels.

Pre-Assessment checks are tests or criteria that implementers can use to determine if they are ready to assess conformance. The intent of specifying these would be to help implementers prepare for conformance testing, not to create a new level of conformance. Examples of pre-assessment checks might be:

The video player used has the ability to display captions and support multiple audio tracks,
All non-text elements on a page have an accessible name.

Editor's note

We continue to explore how the WCAG 2 concept of accessibility-supported fits into proposed conformance models.

End of note

When evaluating the accessibility of content, WCAG 3 requires the outcomes apply to a specific scope. While the scope can be an all content within a digital product, it is usually one or more sub-sets of the whole. Reasons for this include:

Large amounts of content are impractical to evaluate comprehensively using anything beyond automated evaluation of items;
In many cases, content changes frequently, causing evaluation to be accurate only for a specific moment in time;
Some content is more important to the majority of users than other content; and
Content that mostly meets the requirements but has problems can interfere with the user’s ability to complete a process.

WCAG 3 therefore defines two ways to scope content: views and processes. Evaluation is done on one or more complete views or processes, and conformance is determined on the basis of one or more complete views or processes.

Conformance is defined only for processes and views. However, a conformance claim may be made to cover one process and view, a series of processes and views, or multiple related processes and views. All unique steps in a process MUST be represented in the set of views. Views outside of the process MAY also be included in the scope.

Editor's note

We recognize that representative sampling is an important strategy that large and complex sites use to assess accessibility. While it is not addressed within this document at this time, our intent is to later address it within this document or in a separate document before the guidelines reach the Candidate Recommendation stage. We welcome your suggestions and feedback about the best way to incorporate representative sampling in WCAG 3.

End of note

For technology to conform to WCAG 3, the following conformance requirements apply:

Conformance level - Content MUST meet the requirements of the selected conformance level.

Conformance claims are not required. Authors can conform to WCAG 3 without making a claim. The material below describes how to make a conformance claim if that option is chosen.

A conformance claim MUST include the following information:

Date of the claim;
Guidelines title, version and URI W3C Accessibility Guidelines 3.0 at https://www.w3.org/TR/wcag-3.0/
Conformance level satisfied: (bronze, silver, or gold);
A concise description of the views and processes, such as a list of URIs for which the claim is made, including any state changes which lead to a new view; and
The technology including the hardware, software, and assistive technology used to test the claim.

On 12 August 2020, the following 10 views and 2 processes conform to WCAG 3 at a bronze level. Processes were selected because they are the most common activities on the web site and include 4 unique views. The other 6 views are the most commonly used.

Process 1: Order Newsletter (www.example.com, www.example.com/newsletter)
Process 2: Comment on Blog Post (www.example.com, www.example.com/blog, www.example.com/blog (with comment model open)
www.example.com/contact
www.example.com/search_results
www.example.com/recent_news
www.example.com/about
www.example.com/leadership
www.example.com/learn_more

These were tested using Firefox and Chrome on a Windows platform. The assistive technology used included JAWS and Dragon.

Editor's note

We continue to explore how the WCAG 2 concept of conforming alternative versions fit into proposed conformance models.

End of note

This section (with its subsections) provides requirements which must be followed to conform to the specification, meaning it is normative.

Note

Many of the terms defined here have common meanings. When terms appear with a link to the definition, the meaning is as formally defined here. When terms appear without a link to the definition, their meaning is not explicitly related to the formal definition here. These definitions are in progress and may evolve as the document evolves.

End of note

Adequacy

Adequacy is subtle metric, but important to WCAG 3 proposals. Adequacy describes if the formulas being used to process and score the accessibility testing results are using such a small interval that small changes in accessibility do not cause large changes in scoring. Benchmarking Web Accessibility Metrics, Vigo, Lopes, O Connor, Brajnik, Yesilada 2011.

Adjectival Ratings

A system to report evaluation results as a set of human-understandable adjectives.

Assertion

A formal claim of fact, attributed to a person or organization. An attributable and documented statement of fact regarding procedures practiced in the development and maintenance of the content or product to improve accessibility.

Automated evaluation

Evaluation conducted using software tools, typically evaluating code-level features and applying heuristics for other tests.

Automated testing is contrasted with other types of testing that involve human judgement or experience. Semi-automated evaluation allows machines to guide humans to areas that need inspection. The emerging field of testing conducted via machine learning is not included in this definition.

Best Practice

Methods which are not required and meet a higher requirement than methods required to conform to Bronze.

Complexity

Complexity refers to the resources required to accomplish the conformance testing. These could be crawler time, or time for human judgment testing. This would be a useful metric to have to answer the question of how much time WCAG 3 takes to test as compared to WCAG 2. Benchmarking Web Accessibility Metrics, Vigo, Lopes, O Connor, Brajnik, Yesilada 2011.

Conformance

Satisfying all the requirements of the guidelines. Conformance is an important part of following the guidelines even when not making a formal Conformance Claim.

See Conformance.

Deprecate

To declare something outdated and in the process of being phased out, usually in favor of a specified replacement.

Deprecated documents are no longer recommended for use and may cease to exist in the future.

Equity

Equity is the outcome of processes and actions that ensure the spectrum of human reality obtains what is needed to participate, not solely access. As equity relates to WCAG it is about the impact the standards/guidelines have on people with disabilities, along with actually including people with disabilities in the work.

Evaluation

The process of examining content for conformance to these guidelines.

Different approaches to evaluation include automated evaluation, semi-automated evaluation, human evaluation, and user testing.

Functional need

A statement that describes a specific gap in one’s ability, or a specific mismatch between ability and the designed environment or context.

Guideline

High-level, plain-language content used to organize outcomes.

Guidelines provide a high-level, plain-language version of the content for managers, policy makers, individuals who are new to accessibility, and other individuals who need to understand the concepts but not dive into the technical details. They provide an easy-to-understand way of organizing and presenting the outcomes so that non-experts can learn about and understand the concepts. Each guideline includes a unique, descriptive name along with a high-level plain-language summary. Guidelines address functional needs on specific topics, such as contrast, forms, readability, and more. Guidelines group related outcomes and are technology-independent.

Human evaluation

Evaluation conducted by a human, typically to apply human judgement to tests that cannot be fully automatically evaluated.

Human evaluation is contrasted with automated evaluation which is done entirely by machine, though it includes semi-automated evaluation which allows machines to guide humans to areas that need inspection. Human evaluation involves inspection of content features, by contrast with user testing which directly tests the experience of users with content.

Informative

Content provided for information purposes and not required for conformance.

Method

Detailed information, either technology-specific or technology-agnostic, on ways to meet the outcome as well as tests and scoring information.

Normative

Content whose instructions are required for conformance.

Outcome

Result of practices that reduce or eliminate barriers that people with disabilities experience.

See Outcomes.

Point of Regard

The position in rendered content that the user is presumed to be viewing. The dimensions of the point of regard can vary. For example,it can be a two-dimensional area (e.g. content rendered through a two-dimensional graphical viewport), or a point (e.g. a moment during an audio rendering or a cursor position in a graphical rendering), or a range of text (e.g. focused text), or a two-dimensional area (e.g. content rendered through a two-dimensional graphical viewport). The point of regard is almost always within the viewport, but it can exceed the spatial or temporal dimensions of the viewport (see the definition of rendered content for more information about viewport dimensions). The point of regard can also refer to a particular moment in time for content that changes over time (e.g. an audio-only presentation). User agents can determine the point of regard in a number of ways, including based on viewport position in content, keyboard focus, and selection.

Process

A sequence of steps that need to be completed to accomplish an activity / task from end-to-end.

Reliability

The reproducibility and consistency of scores i.e. the extent to which they are the same when evaluations of the same resources are carried out in different contexts (different tools, different people, different goals, different time). This would be particularly useful to ensure that similar results are achieved by different testers. It would also be useful to see if different testers would select the same path or off-path decisions. Representative sampling tests also fit in this category. Benchmarking Web Accessibility Metrics, Vigo, Lopes, O Connor, Brajnik, Yesilada 2011.

Semi-Automated Evaluation

Evaluation conducted using machines to guide humans to areas that need inspection.

Semi-automated evaluation involves components of automated evaluation and human evaluation.

Sensitivity

Sensitivity of a metric is related to the extent that changes in the output of the metric are quantitatively related to changes of the accessibility of the web site being analyzed. This metric is useful for determining if the conformance proposal captures the impact of the severity of accessibility barriers on the final score and if different disabilities are treated equally by the proposal. Benchmarking Web Accessibility Metrics, Vigo, Lopes, O Connor, Brajnik, Yesilada 2011.

Set of Tests

A group of tests that supports a method.

Test

Mechanism to evaluate implementation of a method.

Technique

Technology-specific approach to follow a method.

User need

The end goal a user has when starting a process through digital means.

User testing

Evaluation of content by observation of how users with specific functional needs are able to complete a process and how the content meets the relevant outcomes.

Validity

The extent to which the measurements obtained by a metric reflect the accessibility of the web site to which it is applied. Does the rating that a web site or digital product achieve in any conformance proposal actually reflect the rating that it should get? Benchmarking Web Accessibility Metrics, Vigo, Lopes, O Connor, Brajnik, Yesilada 2011. Accessed on 29 July 2020

WCAG 2	WCAG 3
Success Criteria	Outcomes
Techniques	Methods
Understanding	How-to

W3C Accessibility Guidelines (WCAG) 3.0

Abstract

Status of This Document

1. Introduction

1.1 About WCAG 3

1.2 Section status levels

GuidelinesPlaceholder

Animation and movement

Audio shiftingExploratory

Flashing and strobingExploratory

MotionExploratory

Forms, inputs, and errors

Allow automated entryExploratory

Error associationExploratory

Error identificationExploratory

Error notificationExploratory

Input instructionsExploratory

Input labelsExploratory

Persistent error notificationExploratory

Visible errorExploratory

Processes and task completion

Adequate timeExploratory

Action requiredExploratory

Avoid manipulationExploratory

Go back in processExploratory

Inform at startExploratory

No cognitive testsExploratory

No memorizationExploratory

Optimized processesExploratory

Optional informationExploratory

PreselectionsExploratory

Save progressExploratory

Steps and directionExploratory

Task completion documentationExploratory

Image and media alternatives

AI editableExploratory

Audio alternative in preferred languageExploratory

Audio descriptionsExploratory

CaptionsExploratory

Complex image alternativeExploratory

Context in image alternativeExploratory

Decorative image alternativeExploratory

Descriptive transcriptsExploratory

Finding media alternativesExploratory

Identify autogenerated textExploratory

Image alternativesExploratory

Image roleExploratory

Image typeExploratory

Images-of-Text alternativesExploratory

Non-text alternativesExploratory

Non-verbal cuesExploratory

Persistent figure captionsExploratory

Use of colorExploratory

Use of visual depthExploratory

Use of soundExploratory

Use of spatial audioExploratory

Interactive components

Behavior of controlsExploratory

Change focus with pointer deviceExploratory

Control labelsExploratory

ConventionsExploratory

Consistent labelsExploratory

Control importanceExploratory

Control updatesExploratory

Deceptive controlsExploratory

Distinguishable controlsExploratory

Hover informationExploratory

Interaction indicators contrastExploratory

Input controlExploratory

Name, role, value, stateExploratory

Non-Text contrastExploratory

Notify on changeExploratory

Notify before activationExploratory

Restore focusExploratory

Relevant focusExploratory

Target sizeExploratory

Visual design of controlsExploratory

Input / operation

Comparable keyboard effortExploratory

Consistent keyboard interactionExploratory