As you know, W3C released WCAG 3.0 first draft on the 21st of January 2021. The full form of WCAG also has been changed from web content accessibility guidelines to W3C accessibility guidelines. WCAG 3.0 does not talk web alone and it talks various technologies such as mobile, apps, PDF, ePub, voice input, smart watches, smart
TVs, VR and AR, and so on. unlike WCAG 2.x, WCAG 3.0 aims to be very simple language and cover a broader group of people with disabilities. The guidelines also address related web tools such as user agents (browsers and assistive technologies), content management systems, authoring tools, and testing tools. Working group also wants to update the guidelines on the regular basis in order to keep pace with constant technological change.
WCAG 3.0 is undergoing major changes especially in terms of it’s structure of the guidelines as well as conformance model. In the coming sections, we are going to discuss briefly what changes WCAG 3.0 is undergoing in terms of it’s structure as well as conformance
How are the guidelines are structured
The structure of the WCAG 3.0 is completely changed from WCAG 2.x series. There is no success criterion concept in 3.0 and there are new terms such as outcomes, methods, critical errors, functional categories, you would be hearing.
Guidelines are kind of same as WCAG 2.x series. Guidelines are high-level, plane-language and aimed at non-technical experts for easy understanding. Guidelines are more granular and easier to understand. Guidelines provide the solutions for the accessibility problems. The examples of guidelines are text alternative, clear words, captions, headings and structure, colour contrast. Guidelines are grouping of similar areas to test.
Each guideline contains one or more outcomes. Outcomes are the testable statements. Outcome is similar to success criteria in the 2.x series. Outcomes are measured with scoring but not pass fail. Outcomes are more user-need oriented instead of technology oriented. One of the examples for the outcome is text alternative available.
Functional categories are the group of disabilities. There are 14 disability types as part of functional categories. Each outcome relates to the one or more functional categories. For example, for text alternative available outcome, the functional categories are sensory- vision and visual, cognitive-learning, cognitive-memory, and so on.
Critical errors section is being expected to be revised after receiving the public feedback from the initial draft. As per the initial draft, Critical errors are nothing but a blocker or stopper for the user to access particular features. Critical errors include things like flashing content, audio without pause, keyboard trap. Outcomes may or may not have any critical errors. If outcomes have critical errors, then it would score very low. The scoring information related would be discussed in the next sections of this blog post.
Methods are kind of similar to techniques in the 2.x series. Methods are the ways to meet the outcomes. Outcome may have one or methods. Methods are the areas of concepts to test but methods are not tests though. For example; for text alternatives outcome, the methods are informative images, functional images, decorative images, images of text and so on. in addition, methods vary based on the technology. For example, testing PDF and VR is different from testing HTML.
How-to is similar to the understanding document of 2.x series. It provides the detailed explanation about guidelines. it helps for developers, designers, new to accessibility experts in understanding the guidelines in-detail.
How is WCAG 3.0 tested?
There are 2 types of tests in the WCAG 3.0. one is atomic tests and other one is holistic tests.
Atomic tests are nothing but every element is being tested at the code level. To put it simpler, they are the similar way that we are testing today in the 2.x series. Atomic tests include existing level A, AA, AAA of 2.x series. Atomic tests include both views and process. View is kind of web page. Process is something like flow that user has to follow in order to accomplish the task. Example of process is placing an order in the ecommerce site. Placing order includes the tasks such as searching for the item, adding item to the cart, entering billing and shipping details, receiving the confirmation, and so on. When we complete atomic tests then we reached to bronze conformance level
holistic tests are nothing but testing with real end users who is people with disabilities. Based on the end user feedback, site may reach to silver or gold. we many times see that site claims to be compliant but not usable to people with disabilities, and this is actual problem today. To address this problem, holistic tests should help. There is not much information provided on holistic tests today in the official W3C document, but it is likely to be updated.
Scoring and rating
WCAG 2.x series is kind of more of pass/fail against each success criterion whereas WCAG 3.0 is kind of more of scoring for each outcome. The problem with pass/fail approach of WCAG 2.x is that we fail entire site/view even if one instance of particular checkpoint is not passed, and it is hard to measure like this if the view or site is really compliant or not. WCAG 3.0 is aiming to solve this problem with the new approach called scoring and rating. Let us understand how this scoring works. Outcomes have the methods and methods have the tests to carry out based on the technology. outcomes will be assigned the scoring range from 0(very poor) to 4(excellent). The scoring of outcomes is also dependent on the number of critical errors that are found. If there are critical errors, then scoring goes very low. another important point is that overall scoring must be 3.5 in order to comply the site/view at the bronze level. the overall scoring is depending on the 2 factors. One factor is that average of aggregate outcome must be at least 3.5 rating. Another factor is that each functional category must be at least 3.5 scoring. Please note that the scoring that is being discussed here is for atomic tests whereas the scoring of holistic tests at silver and gold level is not having much information in the W3C document. The below details clearly provide how the scoring is assigned
- Very Poor (0): Any critical errors or less than 50% of related tests pass
- Poor (1): No critical errors, approx. 50% to 79% of related tests pass
- Fair (2): No critical errors, approx. 80% to 89% of related tests pass
- Good (3): No critical errors, approx. 90% to 98% of related tests pass
- Excellent (4): No critical errors, approx. 99% to 100% of related tests pass
Based on the above rating information, site/view is said to be accessible even 99% of the things are accessible but need not to be 100%, and this seems to be good approach I personally feel.
How the conformance model is structured
WCAG 3.0 has 3 levels of conformance, and they are bronze, silver, and gold. WCAG 3.0 does not have any more level A, level Aa, level AAA.
To comply with WCAG 3.0, site/app/any technology meet at least bronze level. The success of atomic tests reaches the site/app/any technology to the bronze level. In order to reach the bronze level, site/app/any technology must need the total score and score within the functional category to be at least 3.5 rating. in addition, there should not be any critical errors in both views and process at the bronze level. Another important aspect is that if the content is compliant to the bronze level, then it does not mean that it met all the requirements, but it is meant that it does not have any critical errors and it is minimally compliant.
Silver and gold
Silver and gold are the next level of conformance. To comply with silver and gold, content must meet minimum bronze level as well as must be usable for the people with disabilities. Holistic tests are needed to perform usability testing with the real users. There is not much information provided on the silver, gold, holistic testing approach in the official W3C document and it is likely to be updated.
Important data points to remember
- WCAG 3.0 is not backward compatible with WCAG 2.X (2.0, 2.1, or 2.2).
- WCAG 3.0 does not supersede WCAG 2.2 and previous versions. It is an alternative set of guidelines.
- WCAG 3.0 will likely not be final until sometime in 2023. More details concerning the scoring mechanism, which is not yet finished, and more outcomes, are an essential part of upcoming “heartbeat” updates, which will occur periodically until WCAG 3.0 becomes a final W3C recommendation.
- There are no more A, AA, AAA levels in WCAG 3.0, so don’t go looking for them. The scoring mechanism replaces them.