WCAG 2.2 Accessibility Quiz

An address form can look complete while still blocking someone who cannot use a mouse, distinguish its error colors, or remember information from an earlier step. Web accessibility concerns whether people can obtain information and finish an activity using the ways they perceive and interact with the web, not just whether a page looks orderly.

The Web Content Accessibility Guidelines (WCAG) 2.2, published by the World Wide Web Consortium (W3C), turn many accessibility needs into testable requirements. Each success criterion addresses a particular outcome under stated conditions. The requirements concern text, images, media, navigation and interactive controls, and are organized around four principles.

Perceivable

Information must be available in a form people can perceive. Examples include meaningful image alternatives, captions and text that remains readable when enlarged.

Operable

People must be able to navigate and use controls. Keyboard access, visible focus and alternatives to difficult pointer movements address different barriers.

Understandable

Content and interaction must be understandable. Identifying languages, explaining errors and avoiding unexpected navigation help people complete tasks.

Robust

Browsers and assistive technologies must be able to interpret the content reliably, including a control's purpose and changing state.

Assistive technologies include screen readers, which present content through speech or braille, and tools that support alternative ways of entering information. They need more than a visual resemblance to a usable interface. Bold cells can look like table headings without telling a screen reader which heading describes each value. Making that relationship programmatically determinable means expressing it so software can identify it, rather than leaving it to visual inference.

Purpose changes the appropriate alternative. A decorative swirl beside a heading can be ignored, while a sales chart that is the only source of important values needs an accessible equivalent of its information. Media has similar distinctions. Captions convey speech and meaningful sounds; audio description conveys important visual events that the soundtrack does not already explain. One does not automatically replace the other.

Several criteria can apply to the same feature. Darkening red error text may improve its contrast, but a red outline alone still does not explain which input failed or what went wrong. Likewise, making a button work from a keyboard does not establish that its focus is visible or that a banner leaves it in view. Accessibility work often requires checking these related requirements separately.

Conformance means satisfying the requirements of the standard. Its levels, A, AA and AAA, are cumulative rather than alternative checklists. Level AA includes Level A requirements; it is not a grade awarded to a person or a percentage of checks passed. An inaccessible required checkout step cannot be compensated for by several accessible product pages.

Automated checks can identify some defects, but human evaluation is needed to judge meaning and interaction. An image alternative can exist yet communicate the wrong information, and a control can be named yet be impossible to operate through a complete task. Evaluation with people with disabilities adds insight into actual usability. Even WCAG conformance does not establish that every disabled person's needs are met.

How to Use This Tool:

The default round selects ten questions from All scenarios, a bank of 38 original text scenarios, with After each answer feedback. Choose a narrower focus when you want to study one subject.

1. Choose Practice focus. Select mixed practice, a principle, an exact criterion level, a role for designers, developers, testers or authors, or Foundation or Applied difficulty. These are alternative focuses, not filters that can be combined. Standards Context groups questions about versions, conformance and evaluation; it is not a fifth WCAG principle.
   A choice such as Level — AA criteria only selects only AA questions. It does not include the Level A requirements needed for AA conformance.
2. Choose Feedback timing. After each answer reveals correctness and the explanation before you advance. At the end records each selection without revealing correctness until the round is complete. Both use the same scoring rules.
3. Select Question count from the available choices. Counts are 1, 3, 5 and 10 where they fit, plus the selected pool's full size. For example, mixed practice offers 38 for the full bank, while Robust practice contains only two questions. Changing focus may change the available counts. If Start practice is disabled, choose an available focus, feedback timing and count, and correct any seed warning.
4. Enter an optional Replay seed under Advanced. Leave it blank for a generated seed, or use up to 128 characters. Reproducing the same questions and option order requires the same seed, focus, count, feedback timing and question-bank version.
5. Select Start practice, then choose one of four answers for each scenario. Selecting an answer records and locks it immediately; there is no separate submit step. Use Next question after each response and View results after the last. A final result requires answering every question.
6. Review Missed answer review before choosing Retry missed, which starts a new round containing only incorrect answers. Replay same seed repeats the current round; Try a new seed makes another seeded round and may repeat questions. A perfect round has no missed-answer retry. Change setup clears the attempt so you can choose another focus.

Interpreting Results:

Practice status shows answered questions during a round and correct answers out of the selected total at completion. The progress percentage measures completion, not correctness. Perfect round means every answer in that attempt matched the authored answer key, not that every WCAG criterion was tested.

Topic performance shows correct and missed counts within sampled principle–level groups, such as Perceivable · AA. Standards context is reported separately, and untested groups are absent rather than awarded a zero or a pass. A longer bar may simply reflect more questions. Review focus identifies a category with the most misses, not a validated measure of your weakest skill or overall proficiency.

The Answer ledger pairs each scenario with your answer, the correct answer, its explanation and source information. Use the explanation to identify the requirement or exception you overlooked. Keep focus, count and feedback timing consistent when comparing practice rounds, and allow for familiarity with repeated questions.

There is no official pass mark or proficiency band. This is not an accessibility audit, conformance certificate, exam qualification or proof of professional competence. Applying a correction to a real website still requires checking the relevant criterion against actual content and interactions, including manual evaluation where needed.

Technical Details:

A success criterion tests a particular outcome, not the overall appearance of a website. First establish whether its covered content or interaction is present, then check its exceptions, then evaluate the requirement. An exception to one criterion does not remove obligations under other criteria.

Criteria and cumulative conformance

Conformance levels accumulate requirements. They cannot be calculated by averaging successful checks or by counting only the criteria carrying the target level's label.

A criterion is satisfied when there is no content that violates it, including when no content falls within its scope. The standard also permits a conforming alternate version where all conditions for that alternative are met. Neither route permits ignoring the remaining conformance requirements.

Conformance concerns full pages, including their responsive variations, and every required page in a complete process. Essential information and functions must be available through accessibility-supported uses of technology, meaning uses that work with the relevant browsers and assistive technologies. Unsupported or nonconforming content must not block access to the rest of the page, including when technologies not relied upon are enabled, disabled or unsupported. The non-interference requirement additionally applies Audio Control, No Keyboard Trap, Three Flashes or Below Threshold, and Pause, Stop, Hide to all page content.

Success criteria and conformance requirements are normative requirements. The Understanding documents and techniques are informative guidance: they explain intent and possible methods without making every listed technique mandatory. A different method can meet a criterion when it delivers the required outcome with the necessary accessibility support.

WCAG 2.2 removed 4.1.1 Parsing. Markup errors can still cause failures under remaining criteria, such as incorrect control names or states under 4.1.2. Removing Parsing does not make all invalid markup acceptable, and an obligation referring to an earlier WCAG version needs to be checked against that version separately.

Meaning, keyboard access and predictable interaction

An accessible name identifies a control to assistive technology. Under 4.1.2, software must be able to identify the control's name and role, set its user-adjustable states, properties and values, and receive notification of changes. A custom checkbox therefore needs more than a drawn checkmark. Under 2.5.3, the accessible name must contain the visible label text, allowing a speech-input user to refer to the words they see. Status messages under 4.1.3 need appropriate roles or properties so they can be presented without moving focus to the message.

Keyboard operation under 2.1.1 covers functionality unless its underlying purpose requires path-dependent input. The separate 2.1.2 requirement prevents keyboard traps and requires instructions for nonstandard exit methods. Focus order under 2.4.3 must preserve meaning and operation, while 2.4.7 requires a visible keyboard focus indicator. Correct focus order does not require every layout to follow an identical visual sequence.

An automatically moving, blinking or scrolling item that runs for more than five seconds alongside other content needs a pause, stop or hide mechanism under 2.2.2, unless the movement is essential. Predictability matters for forms too. Under 3.2.2, changing a setting must not automatically cause a change of context unless the user was advised beforehand. Updating a price in place is not automatically equivalent to submitting a form and navigating elsewhere.

Text contrast, reflow and spacing

Text contrast thresholds depend on the criterion level and whether the text meets the large-scale definition. Each minimum below is inclusive: the actual contrast ratio must reach it before display rounding.

Large-scale text is at least 18 point regular or 14 point bold, or an equivalent size for the relevant script. Eighteen point regular text corresponds to 24 CSS pixels. Cascading Style Sheets (CSS) pixels are page-layout units, not physical screen pixels. Ordinary 16 CSS pixel regular-weight body text does not qualify for the large-text threshold. Logos and certain incidental text, including text in inactive controls, have exceptions. A ratio below the applicable minimum cannot be rounded into a pass.

Reflow under 1.4.10 preserves information and functionality without requiring scrolling in two dimensions at a width equivalent to 320 CSS pixels for vertically scrolling content, or a height equivalent to 256 CSS pixels for horizontally scrolling content. Parts that genuinely require two-dimensional layout, such as a data table or map, can be exceptions. Their presence does not exempt the surrounding article.

Text Spacing under 1.4.12 concerns tolerance of user overrides, not mandatory default typography. With no other style changes, content and functionality must survive the following spacing settings.

The requirement applies to markup-based content that supports these properties. For languages or scripts that do not use a particular property, only the applicable properties need to be considered. Increasing spacing must not make information disappear through clipping or overlap.

The nine additions in WCAG 2.2

All nine additions are represented in the practice bank. Their boundaries differ, even when two criteria address the same feature. The following rules summarize their requirements rather than replace their full definitions and exceptions. A user agent is software that retrieves and presents web content, such as a browser.

For 2.4.11, the notes limit testing of user-movable content to its initial position. They also address user-opened content: a focused control is not considered hidden when the user can reveal it without advancing keyboard focus. Focus Appearance has separate exceptions when the user agent determines an indicator that the author cannot adjust, or when the author has changed neither the indicator nor its background color. Its area requirement does not mandate one outline shape.

The target-size spacing exception uses a 24 CSS pixel diameter circle centered on each undersized target's bounding box. That circle must not intersect another target or the corresponding circle around another undersized target. Other exceptions include a qualifying equivalent control on the same page, inline text links, unmodified user-agent-determined targets, and an essential or legally required presentation. The 44 by 44 CSS pixel threshold belongs to the separate enhanced target-size criterion, not the AA minimum.

A cognitive function test requires mental work such as remembering a password or solving a puzzle. Supporting password-manager entry and paste can provide assistance instead of forcing unaided recall or transcription. That assistance does not remove security requirements. Likewise, an exception for recognizing objects at AA does not mean the same authentication step satisfies AAA or every other applicable criterion.

Practice score and completion

Each correct answer earns one point and each incorrect answer earns zero. There is no partial credit, difficulty weighting, reverse scoring or penalty. Let C be the number of correct answers, N the selected question count, and R the number answered. These are integer question counts satisfying 0 ≤ C ≤ R ≤ N, with N ≥ 1. Possible points equal N, and earned points equal C, ranging from zero to N points. Accuracy and progress have percent units.

Accuracy divides correct answers by all selected questions, not only the questions already answered.

Progress instead measures how much of the selected round has been answered.

The denominator remains the full selected count during an incomplete attempt. Unanswered questions are neither incorrect answers nor missed-retry items. Final review requires all questions to be answered. Accuracy retains its unrounded ratio. Only the visible progress label rounds to the nearest whole percent, with half-percent ties upward. The main completed score remains the exact correct/total count.

Selection is seeded and without replacement within a round. There is no requirement to balance topics in a short sample. The bank contains 16 scenarios assigned to Level A, 13 to AA, four to AAA, and five separate standards-context scenarios. These are question counts, not counts of all WCAG criteria. Several scenarios can concern the same criterion, and a full-bank round still does not cover the whole standard.

Limitations and Privacy:

The question bank is dated 11 September 2026 and identifies the WCAG 2.2 Recommendation of 12 December 2024 as its standards reference. It records an Understanding-document source check of 11 September 2026. Later changes to those explanatory documents do not automatically update the scenarios. Questions and explanations are independently authored, without W3C endorsement.

Scoring happens in the browser, but setup, seed, selected answers and the current view are carried in the page URL, including for unfinished attempts. Someone receiving that link can restore the choices. Saved attempt data in the URL is not encrypted or private progress storage. The URL may enter browser history or be transmitted when a page is loaded or shared. The chart also loads third-party resources, so browser scoring is not a promise of offline operation or zero transmission. Avoid personal or confidential information in a replay seed.

Change setup clears the current attempt and its current URL state, not copies already shared or retained elsewhere. No account-based progress history is provided.

A Saved attempt unavailable warning means the saved attempt could not be restored, for example because it is invalid or incompatible with the question bank. Choose a valid setup and start a new round. If a warning says the attempt could not be saved in the link, reloading or sharing may restore an earlier view rather than the most recent answer.

Worked Examples:

References:

• Web Content Accessibility Guidelines (WCAG) 2.2, W3C, 12 December 2024.
• Understanding Conformance, W3C.
• What's New in WCAG 2.2, W3C.
• Understanding Contrast (Minimum), W3C.
• Understanding Reflow, W3C.
• Understanding Text Spacing, W3C.
• Understanding Focus Appearance, W3C.
• Understanding Target Size (Minimum), W3C.
• Understanding Techniques for WCAG 2.2 Success Criteria, W3C.
• Selecting Web Accessibility Evaluation Tools, W3C, updated 13 May 2024.