Accessibility Label
The accessibilityLabel is the text that assistive technology reads when the component is focused and AMA requires it by tappable elements.
Expectation
- When: The user focuses the component
- Then: The Screen Reader reads out the label
When a component has the accessibility label, it is the first information the screen reader reads when it gains the focus. For instance, with a button, as soon as it receives focus, the screen reader will initially announce its accessibility label. It will then follow up by describing the button's role and state to give the user a comprehensive understanding of the element's function.
This is especially crucial for icon-only buttons, where the control lacks textual labels. In such cases, the accessibility label serves as the primary means of conveying the button's purpose to the user via the screen reader. Without it, the button's function would be unclear, making the app less accessible.
Example
<Pressable
onPress={contactUs}
accessibilityRole="button"
accessibilityLabel="Contact us"
>
Contact us
</Pressable>
No Accessibility Label
The problem
Let's consider the following example:
<Pressable onPress={contactUs} accessibilityRole="button">
Contact us
</Pressable>
When testing the button with both VoiceOver and TalkBack, they both read:
button, Contact us, double-tap to activate
Because the component has no accessibilityLabel
, only the accessibilityRole
is announced; they read the inner text, if any, and in this case: Contact us. Finally, the last part tells the user that the component can be interacted with by performing a double-tap.
What's happen if no text is available?
<Pressable onPress={goBack} accessibilityRole="button">
<SVGIcon />
</Pressable>
When testing the button with both VoiceOver and TalkBack, they both read:
button, double-tap to activate
Here the assistive technology only reads the role and the action that can be performed with the component. So there is a complete lack of helpful information about what we're going to trigger.
Let's fix it
// 1.
<Pressable onPress={contactUs} accessibilityRole="button" accessibilityLabel="Contact US">
Contact us
</Pressable>
// 2.
<Pressable onPress={goBack} accessibilityRole="button" accessibilityLabel="Go back">
<SVGIcon />
</Pressable>
When testing with the assistive technology, this happens:
Contact us button, double-tap to activate Go back button, double-tap to activate
The accessibilityLabel
is announced first, then the role and the action that can be performed at the end.
For this reason, AMA requires that tappable elements have the accessibilityLabel
defined.
All CAPS Accessibility Label
Severity: WarningScreen readers may interpret capital letters as acronyms, misinterpreting content.
Example: ADD TO THE CART
<Pressable accessibilityLabel="ADD TO THE CART">...</Pressable>
This is how the different screen readers handle the uppercase label:
Voice Over | Talkback |
---|---|
A-D-D to the cart | Add to the cart |
In this case, VoiceOver does the spelling of the word ADD
while talkback reads it correctly.
The remaining words are read correctly by both screen readers.
Example: CONTACT US
Voice Over | Talkback |
---|---|
Contact U.S. | Contact U.S. |
The word CONTACT
is read correctly, but both screen readers spell the word US
as it is interpreted as U.S.
for `United States.
A similar issue happens if a sentence contains the word IT, for example.
AMA dev runtime errors
NO_ACCESSIBILITY_LABEL
This error is used when a pressable element has no accessibilityLabel defined.
This rule is mandatory and cannot be turned off!
NO_UPPERCASE_TEXT
This is used when a component has the accessibilityLabel
prop in all caps.
Is it possible to specify a list of allowed all caps accessibility labels, more info here
Related AMA components
External resources
Some helpful resources about accessibility and all caps.