Course progress: 0 / 20 lessons

PRACTICE 9 OF 9 · Controls and navigation

Tooltip and help text

Check whether critical help text is available on focus and not only on hover.

  • Level: Beginner
  • Time: 5-10 minutes
Objective

Recognize when help text must be persistently available or programmatically associated with a control.

PREPARE

Set up your test

  1. Turn on a screen reader you know how to operate.
  2. Navigate by controls or use Tab where the task asks you to.
  3. Listen for purpose and state, not one exact sequence of spoken words.

WORKING EXAMPLE

Establish the expected behavior

Move focus to the field and listen for the help text.

Use the production key, not the test key.

Expected behavior: critical help is available on keyboard focus and through the field description.

INTENTIONAL DEFECT

Find the barrier

The example below deliberately contains one accessibility defect. Skip the defective example.

Move focus to the field. Determine whether the question-mark help is available without a mouse.

Use the production key, not the test key.
Hint

Hover-only content is not available to keyboard-only users.

End of intentional defect.

DIAGNOSE

Choose the primary cause

Why does the defective help text fail?

EXPLANATION

Understand and repair it

Reveal explanation

What happened

The defective field hides important instructions behind a hover-only tooltip. The input has no programmatic description.

Repair

Prefer persistent visible help text for instructions users need to complete a field. Associate it with the input using aria-describedby.

<p id="api-help">Use the production key, not the test key.</p>
<input aria-describedby="api-help">

Retest

  • Focus the field and confirm the help text is available.
  • Confirm the help does not require hover.
  • Confirm any popup help can be dismissed and does not obscure content.

Report it

Example: The API key instruction is available only through a hover tooltip and is not associated with the input.

Standards and support

This exercise describes behavior rather than promising identical speech across screen reader and browser combinations.

Editorially reviewed 20 July 2026.