# Cheat sheet & Interview Q&A — Web Accessibility (A11y)

Source: https://www.skillbyai.com/en/accessibility/a11y-cheat-sheet

> Revise accessibility with a pre-launch checklist, interview questions and answers, and trusted learning resources.

## Quick checklist

Pre-launch accessibility checklist for quick reference.

```markdown
# Accessibility Pre-Launch Checklist

- [ ] Page has unique, descriptive `<title>`
- [ ] One `<h1>` per page, logical heading hierarchy
- [ ] All images have `alt` text (empty if decorative)
- [ ] All form fields have associated `<label>`
- [ ] Buttons and links have descriptive text
- [ ] Color contrast: text ≥ 4.5:1 (AA)
- [ ] Keyboard navigation works (Tab, Enter, Escape)
- [ ] Focus indicator is visible
- [ ] No keyboard traps
- [ ] Skip link to main content
- [ ] Video has captions and transcript
- [ ] No auto-playing sound
- [ ] ARIA used sparingly and correctly
- [ ] aria-live for dynamic updates
- [ ] Tested with NVDA/VoiceOver
- [ ] Axe DevTools: 0 critical issues
```

## Interview Q&A

**Q: What is A11y?** Numeronym for Accessibility. **Q: WCAG levels?** A (basic), AA (minimum legal), AAA (optimal). **Q: Most common issue?** Missing alt text and form labels. **Q: ARIA first rule?** Use semantic HTML first. **Q: Test SR?** NVDA (free), VoiceOver (Mac), JAWS (paid). **Q: Skip links purpose?** Jump past nav to main content. **Q: Announce updates?** aria-live="polite" or aria-live="assertive". **Q: Keyboard essentials?** Tab/Enter/Escape/Arrow work, focus visible, no traps. **Q: Color not enough?** Pair with text/icons. **Q: Without mouse?** Tab/Enter/Arrow/Escape only.

## Resources

**MDN Web Docs**: accessibility guide. **WCAG 2.1**: official guidelines. **WebAIM**: articles + tools. **The A11Y Project**: community. **axe DevTools**: browser extension. **NVDA**: free screen reader.

## Final quiz 1 of 8

Final quiz

**Quiz:** Which WCAG level is the usual legal baseline?

- [ ] A
- [x] AA
- [ ] AAA
- [ ] Level 3

*Answer:* AA. AA is the level most laws and policies require.

## Final quiz 2 of 8

Final quiz

**Quiz:** What is the minimum contrast ratio for normal text at AA?

- [ ] 2:1
- [ ] 3:1
- [x] 4.5:1
- [ ] 7:1

*Answer:* 4.5:1. Normal text needs 4.5:1; large text needs 3:1.

## Final quiz 3 of 8

Final quiz

**Quiz:** Which is the first rule of ARIA?

- [ ] Add role to every element
- [x] Use native semantic HTML when possible
- [ ] Always use aria-hidden
- [ ] Use tabindex above 0

*Answer:* Use native semantic HTML when possible. Native elements already carry roles, states and keyboard support.

## Final quiz 4 of 8

Final quiz

**Quiz:** What should an informative image's alt text do?

- [ ] Repeat the file name
- [ ] Start with "Image of"
- [x] Describe the image's purpose briefly
- [ ] Be left empty

*Answer:* Describe the image's purpose briefly. Alt text conveys purpose; decorative images use alt="".

## Final quiz 5 of 8

Final quiz

**Quiz:** How do you associate a visible label with an input?

- [ ] Use a placeholder only
- [x] Use <label for> matching the input id
- [ ] Use a title attribute
- [ ] Use a bold <span>

*Answer:* Use <label for> matching the input id. A label bound by for/id is announced and enlarges the click target.

## Final quiz 6 of 8

Final quiz

**Quiz:** Which attribute announces dynamic updates politely?

- [x] aria-live="polite"
- [ ] aria-hidden="true"
- [ ] role="presentation"
- [ ] tabindex="-1"

*Answer:* aria-live="polite". Live regions announce content changes without moving focus.

## Final quiz 7 of 8

Final quiz

**Quiz:** What is the purpose of a skip link?

- [ ] Hide navigation permanently
- [x] Let keyboard users jump past repeated content
- [ ] Speed up page load
- [ ] Replace headings

*Answer:* Let keyboard users jump past repeated content. It saves keyboard users from tabbing through the whole navigation.

## Final quiz 8 of 8

Final quiz

**Quiz:** Why combine automated tests with manual testing?

- [x] Automation finds only part of the issues
- [ ] Manual tests are faster
- [ ] Automation is deprecated
- [ ] Screen readers replace tests

*Answer:* Automation finds only part of the issues. Tools catch roughly a third of issues; keyboard and screen reader checks find the rest.
