In 1 minute
- Demonstrate a complete keyboard task with visible focus and identifiable controls.
- Use native semantics, associated labels and messages that explain corrections.
- Combine automated and manual checks and state what still needs validation.
Front-end accessibility in an interview means showing that an interface can be understood and operated in different ways. Knowing attribute names helps, but it does not replace demonstrating a keyboard task, a labeled field or an error that explains how to continue. Your reasoning becomes clearer when each decision addresses a concrete usability barrier.
Use this guide to review a small screen and explain your decisions. The goal is neither memorizing every rule nor claiming conformance after a quick check. For distributing checks throughout a project, consult front-end automated testing. Treat accessibility as expected behavior alongside loading, validation and error recovery.
Demonstrating front-end accessibility in an interview
Choose the task
Choose a complete task, such as locating information and opening its details.
Use the keyboard
Complete it without a mouse and explain where focus appears, how you identify controls and what feedback you receive.
Describe the blocker
If you find a blocker, describe the observed behavior before suggesting attributes. That prevents treating only a visual symptom while the interaction problem remains.
- Define a task and its expected result.
- Check structure, names, keyboard access and feedback.
- Check magnification and meaning without relying only on color.
- Record the problem, impact and reproduction steps.
If you do not know how an assistive technology behaves, state what still needs validation. Keyboard testing does not reproduce every screen reader experience. In an interview, delimiting your evidence is better than claiming a change works for everyone on every browser and tool combination.
Semantic elements and accessible names
Use links for navigation and buttons for actions. Native elements supply semantics and behavior you would otherwise need to rebuild on a div. Headings should represent content organization rather than desired font size. Someone can use that structure to locate a section without reading the whole page. Choosing the right element reduces code and makes intent more predictable.
An accessible name identifies a control to assistive technology. Prefer clear visible text and native associations when available. An icon-only button needs a name, but an unnecessary aria-label may replace visible text with different wording. Check the computed name and maintain consistency between what people see and what their tools announce.
Keyboard access and focus management
Traverse the screen with Tab and Shift+Tab and observe the order without patching it with positive tabindex values. Document order should support a meaningful sequence. Check that the focus indicator is visible and unobscured. Activate controls using their expected keys and confirm you can leave every component.
When opening and closing a modal layer, explain focus entry, containment and return for the chosen behavior. During ordinary result updates, do not automatically move focus to the first item. A status message can communicate changes without interrupting the current position. Distinguish DOM focus, selection and visual highlighting: they can represent different things and need coherent rules.
Event behavior matters too. Review JavaScript events to explain when cancelling a default action helps and when it breaks native interaction. A click handler alone does not replace full keyboard operation.

Form labels and errors
Do
- Associate labels with fields and keep important instructions visible.
- When detecting an error, identify the field and explain the correction. Preserve entered data.
Avoid
- Using placeholders as labels: they disappear during typing and do not provide the same persistent reference.
- Relying only on red borders or a generic message at the top.
This snippet shows a static association between a label, invalid field and message in an address form. It does not implement validation or submission. A dynamic interface also needs decisions about when to validate, how to announce new messages and where focus should go after submission, based on form size and structure.
export function PostalCodeError() {return (<div><label htmlFor="postal-code">Postal code</label><inputid="postal-code"name="postalCode"autoComplete="postal-code"aria-invalid="true"aria-describedby="postal-code-error"/><p id="postal-code-error">Enter a valid postal code.</p></div>);}
ARIA and decisions in a tabs component
ARIA communicates roles, states and relationships; it does not implement behavior. For tabs, discuss the tab set, selected tab and associated panel. The pattern uses arrow navigation within the set while Tab moves onward. Orientation affects the relevant arrows. Distinguish moving focus from activating a panel: these actions may happen together or separately.
Automatic activation can work when panels appear without noticeable delay. If switching requires waiting, consider manual activation with Enter or Space. Roles alone do not complete the widget: relationships, selected state and focus management must agree. For practice without a supplied solution, use keyboard-navigable tabs.
Automated and manual checks
Automated tools find some problems, including certain missing associations or invalid attribute combinations. They cannot decide whether reading order makes sense or an error message helps someone correct a field. Combine automated checks with keyboard use, magnification and assistive reading on defined tasks. Record the environment and steps so someone else can reproduce the assessment.
In React, also track what happens when components unmount and remount; the React questions explain identity. Types can constrain prop combinations, as discussed in TypeScript interviews, but cannot prove the experience. Finish by separating verified issues from suspicions and describing the check still needed for each suspicion.
Frequently asked questions
What accessibility topics should I study for front-end interviews?
Study semantic HTML, accessible names, keyboard access, focus and form feedback. Demonstrate a complete task and explain the limitations of the checks you performed.
How do semantic HTML and ARIA differ?
Native HTML provides semantics and, for many controls, behavior. ARIA describes information for assistive technology but does not add interaction. Prefer native elements when they meet the need.
How do I test an interface with only a keyboard?
Complete a task with Tab, Shift+Tab and control-specific keys. Observe focus, order, activation and ways out of components. Check focus return after closing temporary content.
Does an automated tool guarantee page accessibility?
No. It checks a limited set of conditions. You still need to evaluate tasks, clarity, sequence and different interaction methods. A clean report is not a validated experience.
Practice with real tests
Solve front-end exercises in the browser, with the same tests a reviewer would run.
See the exercises