Your country

Tools that support it use your country for local currency, number formats, units and paper size. Your choice is saved only in this browser.

Type a name or a two-letter code. Use the up and down arrow keys to move through the countries, Enter to choose one and Escape to close.

Web Accessibility (WCAG 2.2)  Module 1 – Foundations: people, barriers and assistive technology

What web accessibility is and why it matters

What web accessibility means, who meets barriers online and why, and how to read the numbers on disability and on the web without claiming too much.

  • Beginner
  • 15 minutes
  • Examples run with Node.js 24.21.0 and quickjs 0.32.0
  • By MySmartCoPilot
Free lesson, Pro answers: 15 minutes
Free: the whole lesson. With a pass: the model answers to its 2 interview questions.

What you will learn

  • Define web accessibility and tell it apart from usability and inclusion
  • Explain the medical and social models of disability and why web teams work from the social model
  • Classify barriers as permanent, temporary or situational, with everyday examples
  • Interpret headline numbers on disability and on the state of the web without overclaiming
What you’ll learn

What you will learn

  • Define web accessibility and tell it apart from usability and inclusion
  • Explain the medical and social models of disability and why web teams work from the social model
  • Classify barriers as permanent, temporary or situational, with everyday examples
  • Interpret headline numbers on disability and on the state of the web without overclaiming
On this page

In a scan of the top one million home pages, 95.9 % failed at least one check of the WCAG: The Web Content Accessibility Guidelines: W3C's standard for accessible web content, made of testable success criteria at levels A, AA and AAA. WCAG 2.2 is the current version and is also published as ISO/IEC 40500. More in “WCAG 2.2 structure, levels and conformance” that software can run, with 56.1 errors on an average page. Each error is a place where a person who is blind, Deaf, has a tremor or reads slowly can get stuck: a photo with no description, a field with no label, text too faint to read.

You will be able to say what web accessibility is and how it differs from usability and inclusion, explain why the barrier sits in the design rather than in the person, sort barriers into permanent, temporary and situational, and read the headline numbers on disability and on the web without claiming more than they show. One small product page, used five different ways, carries the ideas.

Accessibility means disabled people can use the whole site

Picture a bus-ticket site. A blind traveller who uses a screen reader: Software that reads out what is on the screen, or shows it on a braille display, and lets a person move through and operate a page from the keyboard or with touch gestures. NVDA, JAWS, VoiceOver and TalkBack are screen readers. More in “Disabilities, barriers and assistive technology” can search for buses but cannot choose a seat, because the seat map is a grid of unlabelled squares. A Deaf traveller cannot follow the video that explains refunds. A traveller with a tremor keeps hitting the wrong day in a calendar of tiny cells. Each of them can use part of the site, and none of them can buy a ticket.

W3C’s Web Accessibility Initiative describes web accessibility as websites, tools and technologies designed and built so that people with disabilities can use them: they can “perceive, understand, navigate, and interact with the Web”, and they can “contribute to the Web” (Introduction to Web Accessibility). The last verb matters. Accessibility covers posting a review, uploading a document and filling in a form, not only reading. It also covers the whole task: a site where a screen reader user can read every page but cannot pay is not accessible, however many of its pages pass a check.

Two neighbouring ideas are often mixed up with accessibility. W3C keeps them apart, and so should a team, because each asks a different question and is judged against different people (Accessibility, Usability, and Inclusion):

Idea The question it asks Judged against
Accessibility Can disabled people perceive, understand, navigate, interact and contribute? The needs of people with disabilities, tested with WCAG 2.2
Usability Is the product effective, efficient and satisfying? The users and goals the team specifies (ISO 9241-11)
Inclusion Is everyone involved, as far as possible? Diversity of every kind: disability, language, education, devices and connections

The rows overlap. W3C notes a “significant overlap between accessibility and usability”, and much of what helps disabled people makes a site easier for everyone. The overlap is not complete, though: W3C also warns that usability work “often does not sufficiently address the needs of people with disabilities”. A checkout that ten sighted testers finish in a minute can still be impossible with a keyboard. Inclusion is the widest of the three, and accessibility is the part of it with a testable standard: WCAG 2.2, which is also the international standard ISO/IEC 40500.

Checkpoint

Check: What web accessibility covers

  1. Question 1 of 2 On a forum, screen reader users can read every post, but they cannot write a reply, because the reply box has no label. By W3C's description of web accessibility, is the forum accessible?

    Choose one answer.

    Show the answer: Question 1 of 2

    Answer: No: accessibility includes contributing, not only reading

    W3C describes accessibility as disabled people being able to perceive, understand, navigate, interact with and contribute to the web. A forum that can be read but not answered fails the last two for screen reader users.

  2. Question 2 of 2 In a usability test, ten sighted participants with a mouse finish the checkout in under a minute. What does this tell you about the checkout's accessibility?

    Choose one answer.

    Show the answer: Question 2 of 2

    Answer: Very little, because nobody tested it with a keyboard, a screen reader, zoom or captions

    Usability asks whether the specified users reach their goals effectively, efficiently and with satisfaction. If the specified users have no disability, a good result says nothing about keyboard, screen reader, zoom or caption users.

The barrier is in the design, not in the person

Two ways of thinking about disability lead to two different to-do lists. The medical model of disability: A view of disability as a problem in a person's body or mind, to be diagnosed and treated. It describes the person, so it suggests nothing that a website team could change. treats disability as a problem in a person’s body or mind, to be diagnosed and, where possible, treated; for a website it suggests nothing to fix. The social model of disability: A view of disability as the result of barriers in the world around a person, such as a staircase or an unlabelled button, rather than of the impairment alone, so removing the barrier is the fix. puts the disability in the gap between a person and the world built around them: a staircase disables a wheelchair user, and an unlabelled button disables a screen reader user. The World Health Organization’s own description sits close to the social model: disability comes from “the interaction between health conditions” and a person’s surroundings and circumstances (WHO).

The social model in Indian law

The Rights of Persons with Disabilities Act, 2016 defines a person with disability by a long-term impairment that, “in interaction with barriers”, keeps the person from taking part in society on an equal footing, and its definition of a barrier includes communicational, attitudinal and structural factors (RPwD Act, section 2). A website that cannot be used with a screen reader is such a barrier.

Web teams work from the social model because it turns a complaint into a bug. Follow one customer through a payment twice. Watch which item on the stage changes between the attempt that fails and the one that works:

Interactive · step by step One customer, one screen reader, two versions of Pay now Step through a payment that fails and then works. Watch which item changes between the two attempts.
One customer, one screen reader, two versions of Pay now, step 7 of 7. Showing Customer, Pay now (button element), Screen reader (hears: Pay now, button), Order (placed).keysCustomeruses a screen readerPay nowbutton elementScreen readerhears: Pay now, buttonOrderplacedsubmit

The drawing shows the whole scene; its 7 steps are listed below.

Every step as text
  1. The task. A blind customer has filled the basket and wants to pay. Their screen reader reads the page aloud and passes on every key they press.
  2. Tab goes past it. They press Tab to move to the next control. This Pay now is a div with a click handler, which takes no keyboard focus, so Tab jumps past it to the footer.
  3. Read as plain text. Reading line by line, the screen reader finds the words Pay now but no button role. The div has no role at all, so nothing reliably tells the customer it can be pressed.
  4. Stuck. Nothing on the page says there is a control to press, and Tab never lands on it, so the order is not placed. The customer, their skills and their screen reader all worked as they should; the page did not.
  5. One change to the page. The team replaces the div with a button element and changes nothing else. The customer and their software stay exactly as they were.
  6. Tab stops on it. Now Tab stops on Pay now, and the screen reader announces it as a button named Pay now, so the customer knows it can be pressed.
  7. Order placed. Enter presses the button and the order goes through. The barrier was in the design, and changing the design removed it.

Between the fourth step and the seventh, nothing about the customer changed: the same person, the same skills, the same screen reader. Only the page did. That is the practical meaning of the social model. Every barrier has a place in the code or the design, an owner and a fix, and WCAG is written the same way: its success criteria describe properties of content, such as a button having a name and a role, never properties of people.

The same idea works for a whole page. The program below models one product page as the design decisions that matter, then checks it the way five groups of people use the web: by keyboard alone, with a screen reader, zoomed to 400 %, without telling red from green, and without sound. It is a model, not a browser: each part of the page records only the decision its check reads.

One page, five ways of using it, before and after six fixes JavaScript · barrier_tour.mjs
// One product page, five ways of using it. The page is a model: each part records only
// the design decision that a way of using it depends on. The checks list every barrier
// the page puts in a person's way, then run again on the same page after six fixes.

const before = {
  photo: { alt: null }, // an <img> without an alt attribute
  size: { name: null }, // a <select> with no label
  pay: { element: 'div', focusable: false }, // a div with a click handler
  required: { shownBy: ['colour'] }, // required fields marked in red only
  layout: { minWidth: 1280 }, // the narrowest width the layout fits, in CSS px
  video: { captions: false }, // a product video with speech
};

const after = {
  photo: { alt: 'Steel water bottle, 750 ml, matt blue' },
  size: { name: 'Size' },
  pay: { element: 'button', focusable: true },
  required: { shownBy: ['colour', 'text'] },
  layout: { minWidth: 320 },
  video: { captions: true },
};

// A way of using the page, and what fails it: [part, WCAG criterion, what happens].
const ways = [
  ['Keyboard only', (p) => [!p.pay.focusable && ['pay', '2.1.1', 'Tab never reaches "Pay now"']]],
  ['Screen reader', (p) => [
    p.photo.alt === null && ['photo', '1.1.1', 'the photo has no text alternative'],
    p.size.name === null && ['size', '4.1.2', 'the size list has no name'],
    p.pay.element !== 'button' && ['pay', '4.1.2', '"Pay now" is read as text, not a button'],
  ]],
  ['Zoom to 400 %', (p) => [p.layout.minWidth > 320 && ['layout', '1.4.10', `the page needs ${p.layout.minWidth} px, so it scrolls sideways`]]],
  ['Red looks like green', (p) => [!p.required.shownBy.includes('text') && ['required', '1.4.1', 'required fields are marked by colour alone']]],
  ['No sound', (p) => [!p.video.captions && ['video', '1.2.2', 'the video has no captions']]],
];

function tour(label, page) {
  console.log(label);
  let barriers = 0;
  const blocked = new Set();
  for (const [way, check] of ways) {
    for (const found of check(page).filter(Boolean)) {
      const [part, sc, what] = found;
      console.log(`  ${way.padEnd(21)}${part.padEnd(10)}${sc.padEnd(8)}${what}`);
      barriers += 1;
      blocked.add(way);
    }
  }
  console.log(`  ${barriers} barrier${barriers === 1 ? '' : 's'}, in ${blocked.size} of ${ways.length} ways of using the page`);
  console.log('');
}

tour('Before the fixes', before);
tour('After the fixes', after);

Output

Before the fixes
  Keyboard only        pay       2.1.1   Tab never reaches "Pay now"
  Screen reader        photo     1.1.1   the photo has no text alternative
  Screen reader        size      4.1.2   the size list has no name
  Screen reader        pay       4.1.2   "Pay now" is read as text, not a button
  Zoom to 400 %        layout    1.4.10  the page needs 1280 px, so it scrolls sideways
  Red looks like green required  1.4.1   required fields are marked by colour alone
  No sound             video     1.2.2   the video has no captions
  7 barriers, in 5 of 5 ways of using the page

After the fixes
  0 barriers, in 0 of 5 ways of using the page

Recorded with Node.js 24.21.0 on macOS 26 arm64. To run it yourself: mise exec node@24.21.0 -- node barrier_tour.mjs

Seven barriers, and every one of the five ways of using the page meets at least one of them. The fixes are ordinary work: an alt attribute, a label, a button element, the word “required” next to the red marker, a layout that can narrow, and a captions file. After them the same five checks find nothing.

Undo one fix

Press Edit, set the photo’s alt in after back to null and run the program again. One row comes back, under Screen reader, and the other ways of using the page still meet no barrier.

Each barrier belongs to one decision, so each one can be found, assigned and fixed like any other defect.

Key idea

A barrier is a property of the design, not of the person: change the page and the same person, with the same assistive technology, gets through.

Checkpoint

Check: The barrier is in the design

  1. Question 1 of 2 In the stepped payment, what changed between the attempt where the customer was stuck and the one where the order was placed?

    Choose one answer.

    Show the answer: Question 1 of 2

    Answer: Only the page: the div became a button element

    The person and their assistive technology never changed. Replacing the div with a button gave Pay now keyboard focus and a role, and that alone removed the barrier.

  2. Question 2 of 2 A Deaf customer writes that your product videos have no captions. Under the social model, what is the right response?

    Choose one answer.

    Show the answer: Question 2 of 2

    Answer: Treat it as a defect of the product, add captions and fix the step of publishing that let videos out without them

    The social model locates the barrier in the design: here, a video published without captions. The team owns it, fixes it, and changes how videos are published so the barrier does not come back.

Barriers can last a lifetime, a few weeks or a few minutes

The tiny “x” that closes a pop-up is hard to hit for a person with a tremor, for someone whose wrist is in a cast, and for anyone tapping it on a moving bus. The barrier is the same in all three cases. How long the person meets it is not:

  • Permanent: a long-term impairment, such as blindness, deafness, a limb difference or a learning disability. The person meets the barrier on every visit.
  • Temporary: an impairment that passes. W3C gives accidents, surgery and medication as causes of temporary impairments (Diverse Abilities and Barriers): a broken arm, eyes dilated after an eye test, a concussion.
  • Situational: a situational limitation: A constraint that comes from a person's surroundings or circumstances rather than from an impairment, such as glare, a noisy place or a moving vehicle. It ends when the situation does. comes from the surroundings or the circumstances, such as noise, glare from the sun or the cost of mobile data.

The table puts one example of each against four kinds of ability. Read a row across: the need stays the same while its cause and its length change.

Ability Permanent Temporary Situational
Hearing Deaf since birth An ear infection A loud railway platform
Seeing Blind, or low vision Pupils dilated after an eye test Glare on a phone in the midday sun
Moving A tremor, or one arm A wrist in a cast Holding a child on a moving bus
Thinking A learning disability A concussion Constant interruptions

Two cautions keep this honest. First, a situational limitation is not a disability. The person on the bus can get off; the person with a tremor cannot, and WCAG’s requirements exist for the second. Second, the overlap is a bonus, not the reason. W3C lists people using small screens, older people and people on slow connections among those who also gain from accessible design, but a design is accessible when disabled people can use it, whoever else benefits.

Checkpoint

Check: Permanent, temporary and situational barriers

  1. Question 1 of 2 After an eye test, a customer's pupils stay dilated for a few hours, and the small grey text of a form becomes unreadable. Which kind of barrier is this?

    Choose one answer.

    Show the answer: Question 1 of 2

    Answer: Temporary

    Dilated pupils are a temporary impairment. The same small grey text would be a permanent barrier for someone with low vision and a situational one in bright sun, which is why one fix helps all three.

  2. Question 2 of 2 A colleague says, "Everyone is disabled sometimes, for example in bright sun, so let us design for situations instead of for disabled people." What is wrong with that?

    Choose one answer.

    Show the answer: Question 2 of 2

    Answer: A situational limitation ends with the situation, while a disabled person meets the barrier on every visit, and WCAG's requirements exist for them

    Designing for situations helps many people, but it can miss needs that no situation reproduces, such as a screen reader user's need for names and roles. Accessible means disabled people can use it, whoever else gains.

Headline numbers measure different things

Accessibility arguments lean on two kinds of number: how many people have a disability, and how much of the web fails them. Both are easy to quote and easy to stretch, so look at what each one actually counted.

The World Health Organization estimates that 1.3 billion people, 16 % of the world’s population, “experience a significant disability” (WHO). It is an estimate of significant disability across the world, not a count of anyone’s users, and it says nothing about how many of those people visit a particular website. Analytics cannot fill the gap either, because browsers do not tell a website which assistive technology a visitor uses.

India's census count

Census 2011 counted 2.68 crore persons with disabilities in India, 2.21 % of the population (DEPwD). The census asked its own questions in its own way, and the RPwD Act of 2016, which lists 21 specified disabilities, came after it. Never set 2.21 % against WHO’s 16 % as if one of them were wrong: they count different things.

For the state of the web, WebAIM runs software checks on the home pages of the top one million websites. The program below lays out its results for the six most common kinds of failure, adds the WCAG criterion each one fails and who it stops first, and works out three numbers that the published summary implies:

The six most common failures on a million home pages JavaScript · webaim_six.mjs
// The six most common failure types in the WebAIM Million: the top 1,000,000 home pages,
// scanned by software. Each row: the type, the share of home pages that had it, the
// WCAG criterion that axe-core 4.13.0 tags its matching rule with (its rule for empty
// links carries 4.1.2 as well), and who it stops first.
const types = [
  ['Low contrast text', 83.9, '1.4.3', 'people with low vision'],
  ['Images without alternative text', 53.1, '1.1.1', 'screen reader users'],
  ['Form fields without labels', 51.0, '4.1.2', 'screen reader and voice users'],
  ['Empty links', 46.3, '2.4.4', 'screen reader users'],
  ['Empty buttons', 30.6, '4.1.2', 'screen reader and voice users'],
  ['No document language', 13.5, '3.1.1', 'screen reader users'],
];
const pages = 1000000;
const failingShare = 95.9; // % of home pages with at least one detected failure
const errorsPerPage = 56.1; // detected errors on an average home page
const sixTypesShare = 96; // % of all detected errors that are one of the six types

console.log('Failure type                      Pages   WCAG   Stops first');
for (const [type, share, sc, who] of types) {
  console.log(`${type.padEnd(34)}${share.toFixed(1).padStart(5)} %  ${sc.padEnd(7)}${who}`);
}

// 41000 -> "41,000" (written out, so the program prints the same everywhere).
const grouped = (n) => String(n).replace(/\B(?=(\d{3})+(?!\d))/g, ',');

const shares = types.reduce((sum, t) => sum + t[1], 0);
const clean = 100 - failingShare;
console.log('');
console.log(`The six shares add up to ${shares.toFixed(1)} % of home pages, more than the ${failingShare} % that fail:`);
console.log(`a failing home page has ${(shares / failingShare).toFixed(1)} of the six types on average.`);
console.log(`Errors per page: ${errorsPerPage}, of which ${(errorsPerPage * sixTypesShare / 100).toFixed(1)} are one of the six types.`);
console.log(`Home pages with no detected failure: ${clean.toFixed(1)} %, about ${grouped(Math.round(pages * clean / 100))} of ${grouped(pages)}.`);
console.log('A scan finds only some kinds of failure, so fewer pages than that conform.');

Output

Failure type                      Pages   WCAG   Stops first
Low contrast text                  83.9 %  1.4.3  people with low vision
Images without alternative text    53.1 %  1.1.1  screen reader users
Form fields without labels         51.0 %  4.1.2  screen reader and voice users
Empty links                        46.3 %  2.4.4  screen reader users
Empty buttons                      30.6 %  4.1.2  screen reader and voice users
No document language               13.5 %  3.1.1  screen reader users

The six shares add up to 278.4 % of home pages, more than the 95.9 % that fail:
a failing home page has 2.9 of the six types on average.
Errors per page: 56.1, of which 53.9 are one of the six types.
Home pages with no detected failure: 4.1 %, about 41,000 of 1,000,000.
A scan finds only some kinds of failure, so fewer pages than that conform.

Recorded with Node.js 24.21.0 on macOS 26 arm64. To run it yourself: mise exec node@24.21.0 -- node webaim_six.mjs

Read the last two lines first. 4.1 % of home pages had no failure the software could find, and WebAIM itself says that full conformance is “certainly lower than 4.1%”, because “not all conformance failures can be automatically detected” (WebAIM). The scan covers home pages only, and it cannot say how many people each failure stopped. What it does show is that most failures are the same few, that a failing page has nearly three of them at once, and that five of the six are fixed by adding a missing attribute or a missing word.

Three habits keep claims honest:

  • Quote what was measured: “detected failures on home pages”, not “inaccessible websites”.
  • Never present a clean automated scan as proof that a page is accessible.
  • Claim a benefit only where there is evidence for it. Accessible design helps many people, but it is no promise of better search rankings.
Website Accessibility Checker (WCAG 2.2) Run automated WCAG checks on a page of your own, then list what the scan could not judge.

Checkpoint

Check: What the headline numbers measure

  1. Question 1 of 3 What does "95.9 % of home pages had detected WCAG failures" measure?

    Choose one answer.

    Show the answer: Question 1 of 3

    Answer: The share of the top million home pages on which software found at least one failure

    WebAIM scans one page per site, the home page, for the top million sites, with software. "Detected" is the important word: failures that need a person to judge them are not in the number.

  2. Question 2 of 3 A slide shows a national census that counted 2.21 % of the population as disabled next to WHO's estimate of 16 % worldwide, and says one of them must be wrong. What is the problem with the slide?

    Choose one answer.

    Show the answer: Question 2 of 3

    Answer: The two numbers measure different things, with different questions, definitions and methods

    Quote each number with what it measured. Neither tells you how many of your own users are disabled, and your analytics cannot tell you either.

  3. Question 3 of 3 An automated scan of your home page reports no failures. What can you honestly claim?

    Choose one answer.

    Show the answer: Question 3 of 3

    Answer: That the tool found no failure it can detect on that page

    A clean scan is useful and limited. It removes the detectable failures, and it leaves keyboard paths, the meaning of text alternatives and whole tasks for a person to test.

Practise: the four ideas

Ten questions cover the lesson, from the definition to the WebAIM figures. Each one picks the idea where your record is thinnest, and a right answer brings a harder question on that idea next time:

Practice that adapts to you

Practise this lesson

10 questions about 4 ideas of this lesson. Each answer is explained at once, and what you get right is kept in this browser.

  1. Question 1 of 10 Which list matches W3C's description of what web accessibility lets disabled people do?

    Choose one answer.

    Show the answer: Question 1 of 10

    Answer: Perceive, understand, navigate, interact and contribute

    W3C describes accessibility as people with disabilities being able to perceive, understand, navigate, and interact with the web, and contribute to it.

  2. Question 2 of 10 A team's usability scores are excellent, yet a blind tester cannot submit its sign-up form. Which statement fits W3C's view?

    Choose one answer.

    Show the answer: Question 2 of 10

    Answer: Accessibility and usability overlap but are not the same; usability work often misses disabled people's needs

    Usability is judged against the specified users. Unless disabled people are among them, a good usability result leaves accessibility untested.

  3. Question 3 of 10 Which model of disability places it in the gap between a person and the design around them?

    Choose one answer.

    Show the answer: Question 3 of 10

    Answer: The social model

    The social model puts the barrier in the environment and the design, which is why it gives web teams something to fix.

  4. Question 4 of 10 In the barrier tour, the screen reader meets three barriers on the page before the fixes. Which fix removes the barrier on the size list?

    Choose one answer.

    Show the answer: Question 4 of 10

    Answer: Give the list a name, for example with a label element that says Size

    Each barrier in the tour belongs to one decision, so each fix removes exactly one row: the label removes the size row, and the other rows need their own fixes.

  5. Question 5 of 10 Why are WCAG's success criteria written as properties of content, such as "a button has a name", rather than as properties of users?

    Choose one answer.

    Show the answer: Question 5 of 10

    Answer: Because the barrier and its fix are in the content, so each criterion can be tested on the page

    A criterion about content can be tested by anyone with the page in front of them, and a failure names something the team can change.

  6. Question 6 of 10 Tapping a small button accurately while standing on a moving bus is which kind of barrier?

    Choose one answer.

    Show the answer: Question 6 of 10

    Answer: Situational

    A moving bus is a situational limitation. A tremor would make the same button a permanent barrier, and a wrist in a cast a temporary one.

  7. Question 7 of 10 Which change helps a person with a tremor, a person with a wrist in a cast and a person tapping on a moving bus, all at once?

    Choose one answer.

    Show the answer: Question 7 of 10

    Answer: Larger targets with space between them

    The need is the same in all three durations, so one design fix meets it for all of them.

  8. Question 8 of 10 WebAIM found detected failures on 95.9 % of 1,000,000 home pages. About how many of those home pages had no detected failure?

    Type a number.

    A hint

    Work out the share that had no detected failure first.

    Show the answer: Question 8 of 10

    Answer: 41000 home pages (anything from 40900 to 41100 counts)

    100 % − 95.9 % = 4.1 %, and 4.1 % of 1,000,000 is 41,000, the number the program printed. Real conformance is lower still, because a scan detects only some kinds of failure.

  9. Question 9 of 10 The shares of home pages with each of the six most common failure types add up to 278.4 %, and 95.9 % of home pages fail. On average, how many of the six types does a failing home page have?

    Type a number.

    A hint

    Every page counted in a share is one of the failing pages.

    Show the answer: Question 9 of 10

    Answer: 2.9 types (anything from 2.85 to 2.95 counts)

    278.4 ÷ 95.9 = 2.9, as the program printed. Shares above 100 % mean the types overlap: most failing pages have several of them at once.

  10. Question 10 of 10 Which sentence reports the WebAIM result honestly?

    Choose one answer.

    Show the answer: Question 10 of 10

    Answer: Software found WCAG failures on 95.9 % of the top million home pages.

    Quote what was measured, how and where. Each of the wrong choices turns a count of detected failures into a claim about people.

Interview questions

Fresher · What is web accessibility, and how is it different from usability?

Web accessibility means that people with disabilities can use a website or app as fully as anyone else: perceive its content, understand it, navigate it, operate its controls and contribute, by posting, uploading or filling in forms. Usability asks whether a product is effective, efficient and satisfying for the users it was designed for, and those users are often assumed to have no disability. The two overlap a lot, but a usable site can still be inaccessible: a checkout that most testers finish quickly may have a pay button that no keyboard can reach. Accessibility is judged against the needs of disabled people, and it has a testable standard, WCAG 2.2, also published as ISO/IEC 40500.

Fresher · What is the social model of disability, and why does it matter to a web team?

The medical model sees disability as a problem in a person’s body, to be treated. The social model sees it in the gap between a person and the world built around them: an unlabelled button disables a screen reader user the way a staircase disables a wheelchair user. For a web team, the social model turns a complaint into a defect with a place in the code, an owner and a fix. WCAG is written the same way: every success criterion describes something about the content, such as a button having a name and a role, never something about the user. So the question is never what is wrong with the user, but what in the page stops them.

Two harder questions of this module start from this lesson’s numbers:

Model answers open with any pass: ₹79 for 1 week ($2.99), one payment that never renews.

  1. Fresher to mid

    A product manager says accessibility can wait, because "only about 2 % of people are disabled" and the analytics show no disabled users at all. How do you answer, using numbers you can defend?

    Model answer, Pro. Model answers open with any pass; prices are in the notice on this page. First step and follow-ups

  2. Mid to senior

    An automated scan of your home page reports zero accessibility errors, and the marketing team wants to announce that the site is accessible. What do you tell them, and what would you check before anything is said in public?

    Model answer, Pro. First step and follow-ups

Key takeaways

  • Web accessibility means disabled people can perceive, understand, navigate, interact with and contribute to a site, across whole tasks; it overlaps with usability and inclusion but is judged against disabled people’s needs.
  • The social model places the barrier in the design: change the page, and the same person with the same assistive technology gets through.
  • Barriers can be permanent, temporary or situational, but WCAG’s requirements exist for disabled people, who cannot step out of the situation.
  • WHO’s 16 % is a global estimate of significant disability; a national census measures something else, and neither tells you how many of your users are disabled.
  • WebAIM found detected failures on 95.9 % of the top million home pages; a clean scan shows only that software found nothing, never that a page is accessible.

Terms in this lesson

medical model of disability
A view of disability as a problem in a person's body or mind, to be diagnosed and treated. It describes the person, so it suggests nothing that a website team could change.
situational limitation
A constraint that comes from a person's surroundings or circumstances rather than from an impairment, such as glare, a noisy place or a moving vehicle. It ends when the situation does.
social model of disability
A view of disability as the result of barriers in the world around a person, such as a staircase or an unlabelled button, rather than of the impairment alone, so removing the barrier is the fix.

References

Related tools

Report a problem with this lesson

Quick answers and tool search

Type to search tools or to get a quick answer, for example 18% of 2500. Use the up and down arrow keys to move through the results, Enter to choose, and Escape to close.