Web Fundamentals · Lesson 10 of 10 · 4 min
Accessibility Testing, WCAG & Interview Prep
This is how you prove you don't just talk about accessibility, you ship it.
After this lesson, you'll be able to articulate a professional-grade accessibility testing strategy in an interview. You will know what to test, how to test it, and how to explain the limitations of each method.
Frame your answer around WCAG levels
The Web Content Accessibility Guidelines (WCAG) are the standard. You don't need to recite them, but you must know the structure to sound credible.
WCAG is built on four principles (POUR): content must be Perceivable, Operable, Understandable, and Robust.
Under these principles are success criteria, which have three conformance levels:
- A: The bare minimum. Failing this means you're blocking users.
- AA: The industry and legal standard. This is your target.
- AAA: The highest level. Often not practical for an entire application, but good to aim for on specific, simple components.
In an interview, state that you build and test to WCAG 2.1 Level AA. This is the correct answer. It shows you understand the common requirement and the trade-offs.
The testing workflow: Automated, Manual, Assistive Tech
A robust accessibility strategy has layers. Relying on one method is a common mistake that signals inexperience. You need to describe a process that combines automation for breadth and manual testing for depth.
-
Automated Scans: This is your first line of defense. Tools like
axe-coreor Lighthouse scan for programmatic issues: missingaltattributes, bad color contrast, invalid ARIA properties. They are fast, run in CI, and catch 20-50% of total issues. -
Manual Keyboard Testing: This is non-negotiable. Unplug your mouse and navigate your entire feature using only Tab, Shift+Tab, Enter, Space, and arrow keys. Can you reach everything? Is the focus order logical? Can you operate every control?
-
Screen Reader Testing: This is the only way to understand the experience for blind and low-vision users. Use a real screen reader like VoiceOver (macOS), NVDA (Windows), or TalkBack (Android). Does the content make sense when read aloud? Are interactions announced clearly?
This workflow catches issues early and ensures you're not just checking boxes but building a usable product.
The trap: Over-relying on automated tools. An interviewer might ask about their limitations. Your answer should be that they can't verify context or usability. An automated tool can see an alt attribute exists, but it can't tell you if the text is meaningful. It can't tell you if your tab order is confusing.
Set up automated testing in your pipeline
Show you can put theory into practice by describing how to integrate axe-core into an end-to-end testing suite like Playwright or Cypress. This demonstrates a "shift-left" mentality, where you catch bugs early.
This test navigates to a page, runs an axe-core scan against WCAG 2.1 AA criteria, and fails if any violations are found.
// tests/accessibility.spec.ts
import { test, expect } from '@playwright/test';
import AxeBuilder from '@axe-core/playwright';
test.describe('Homepage Accessibility', () => {
test('should not have any automatically detectable accessibility issues', async ({ page }) => {
await page.goto('/');
const accessibilityScanResults = await new AxeBuilder({ page })
.withTags(['wcag2a', 'wcag2aa', 'wcag21a', 'wcag21aa'])
.analyze();
expect(accessibilityScanResults.violations).toEqual([]);
});
});
This is the baseline. It's simple, effective, and belongs in every project's CI pipeline.
What to look for in manual tests
This is where you demonstrate seniority. Automated tests find the easy stuff. Manual and screen reader testing is where you find the critical usability bugs that automation misses.
When asked what manual testing uncovers, be specific.
Automated tools are a great start, but they can't catch issues with interaction or logic. For example, in a modal dialog, I manually test for three things:
- Focus Management: Does focus move into the modal when it opens? Is it trapped inside, so tabbing doesn't escape to the page behind it?
- Keyboard Operability: Can I close the modal with the
Escapekey? Can I operate all controls inside it?- Focus Restoration: When the modal closes, does focus return to the element that opened it?
An automated tool can't verify this entire flow. You have to test it as a user would.
Other key areas for manual and screen reader testing:
- Logical Focus Order: Does the focus follow a predictable path through the page?
- Meaningful Announcements: For a custom component, does a screen reader announce its role and state correctly? When a filter is applied, does an
aria-liveregion announce that the results have been updated? - Skip Links: Is there a "Skip to main content" link for keyboard users to bypass navigation?
Having a few concrete examples like these proves you've moved beyond automated reports and into the real-world challenges of accessibility.
Try it
Pick a complex interactive component from a site you use, like a date picker or a multi-select dropdown. Spend 15 minutes trying to use it with only your keyboard and a screen reader (VoiceOver or NVDA). Document one thing that works well and one thing that is broken or confusing.
Next: You've completed the course. It's time to apply this knowledge in a realistic setting with a mock frontend interview.
Lock it in — apply this lesson under interview conditions.