Notice: This is a work in progress. Please submit feedback or suggestions.

Test: ARIA alertdialog role must use document mode

Screen Reader support level: supported

Voice Control support level: not applicable (or no 'MUST' expectations)

On this page

About this test

This test ensures the alertdialog role does not invoke application mode in screen readers. The default mode (document) must be used. ARIA 1.0 had language for the alertdialog role that could suggest that application mode must be invoked. However, that language was clarified in aria 1.1.

Run this test and submit your results

Age of results

Results in this test range from a year ago to 3 years ago. See detailed information about test dates and versions for more information.

Test HTML

open the test page
<!DOCTYPE html>
<html lang="en">
<head>
    <title>ARIA alertdialog role document mode test</title>
</head>
<body>
<button>Go to Dialog</button>
<div role="alertdialog" aria-label="test">
    <h1 tabindex="-1">Test Dialog</h1>
    <p>paragraph before</p>
    <p id="target">target</p>
    <p>paragraph after</p>
</div>

<script>
    /** It is necessary to send focus to an element inside of the dialog in order for some screen readers (NVDA) to read its contents. **/
    document.querySelector('button').addEventListener("click", function(){
        document.querySelector('h1').focus();
    });
</script>
</body>
</html>

Summary of Screen Reader support by expectation

alertdialog role
ExpectationJAWSNarratorNVDAOrcaTalkBackVoiceOver (iOS)VoiceOver (macOS)
ChromeIEFirefoxEdgeChromeFirefoxFirefoxChromeSafariSafari
MUST NOT automatically switch away from reading modesupportedsupportedsupportedsupportedsupportedsupportedsupportedsupportedsupportedsupported

* means that some support is hidden behind settings

Summary of Voice Control support by expectation

alertdialog role
ExpectationDragon Naturally SpeakingVoice Access (Android)Voice Control (iOS)Voice Control (MacOS)Windows Speech Recognition
ChromeChromeSafariSafariChrome
Not applicable

* means that some support is hidden behind settings

Detailed support results by expectation

Expectation for the alertdialog role: NOT automatically switch away from reading mode

This expectation is from the alertdialog role feature.

Strength of this expectation for different types of assistive technologies:

  • Screen Readers: MUST
  • Voice Control: NA

Grading method:

Just one of the listed commands must pass for the expectation to pass.

ATBrowserHas SupportOutput
Dragon Naturally SpeakingChromenot applicable-
JAWSChromeyes
  • Result: (pass)
    • Test Case: Use Down arrow (Read next item) to navigate forward to `*[role="alertdialog"]`
      1. Launch JAWS and Chrome.
      2. Navigate to the test page.
      3. Find the target element(s) that you will test against. Identify all elements that match this selector:*[role="alertdialog"]
        • If multiple elements match the selector, repeat this test for all instances. However, choose a single instance to report against. If you feel that the selector should be more specific, please open a GitHub Issue.
      4. Position and configure the screen reader so that the following conditions are met
        • Virtual focus is: before target
        • Keyboard focus is: before target
        • Mode is: auto (do not explicitly change the mode)
      5. Issue the command: Down arrow (Read next item)
        • After issuing the command, virtual focus should be: target
      6. Record results for the relevant expectations
    • Output: target
JAWSIEyes
  • Result: (pass)
    • Test Case: Use Down arrow (Read next item) to navigate forward to `*[role="alertdialog"]`
      1. Launch JAWS and IE.
      2. Navigate to the test page.
      3. Find the target element(s) that you will test against. Identify all elements that match this selector:*[role="alertdialog"]
        • If multiple elements match the selector, repeat this test for all instances. However, choose a single instance to report against. If you feel that the selector should be more specific, please open a GitHub Issue.
      4. Position and configure the screen reader so that the following conditions are met
        • Virtual focus is: before target
        • Keyboard focus is: before target
        • Mode is: auto (do not explicitly change the mode)
      5. Issue the command: Down arrow (Read next item)
        • After issuing the command, virtual focus should be: target
      6. Record results for the relevant expectations
    • Output: target
JAWSFirefoxyes
  • Result: (pass)
    • Test Case: Use Down arrow (Read next item) to navigate forward to `*[role="alertdialog"]`
      1. Launch JAWS and Firefox.
      2. Navigate to the test page.
      3. Find the target element(s) that you will test against. Identify all elements that match this selector:*[role="alertdialog"]
        • If multiple elements match the selector, repeat this test for all instances. However, choose a single instance to report against. If you feel that the selector should be more specific, please open a GitHub Issue.
      4. Position and configure the screen reader so that the following conditions are met
        • Virtual focus is: before target
        • Keyboard focus is: before target
        • Mode is: auto (do not explicitly change the mode)
      5. Issue the command: Down arrow (Read next item)
        • After issuing the command, virtual focus should be: target
      6. Record results for the relevant expectations
    • Output: target
NarratorEdgeyes
  • Result: (pass)
    • Test Case: Use Down arrow (Read next item (scan mode)) to navigate forward to `*[role="alertdialog"]`
      1. Launch Narrator and Edge.
      2. Navigate to the test page.
      3. Find the target element(s) that you will test against. Identify all elements that match this selector:*[role="alertdialog"]
        • If multiple elements match the selector, repeat this test for all instances. However, choose a single instance to report against. If you feel that the selector should be more specific, please open a GitHub Issue.
      4. Position and configure the screen reader so that the following conditions are met
        • Virtual focus is: before target
        • Keyboard focus is: before target
        • Mode is: auto (do not explicitly change the mode)
      5. Issue the command: Down arrow (Read next item (scan mode))
        • After issuing the command, virtual focus should be: target
      6. Record results for the relevant expectations
    • Output: target
NVDAChromeyes
  • Result: (pass)
    • Test Case: Use down arrow (Read next item) to navigate forward to `*[role="alertdialog"]`
      1. Launch NVDA and Chrome.
      2. Navigate to the test page.
      3. Find the target element(s) that you will test against. Identify all elements that match this selector:*[role="alertdialog"]
        • If multiple elements match the selector, repeat this test for all instances. However, choose a single instance to report against. If you feel that the selector should be more specific, please open a GitHub Issue.
      4. Position and configure the screen reader so that the following conditions are met
        • Virtual focus is: before target
        • Keyboard focus is: before target
        • Mode is: auto (do not explicitly change the mode)
      5. Issue the command: down arrow (Read next item)
        • After issuing the command, virtual focus should be: target
      6. Record results for the relevant expectations
    • Output: target
NVDAFirefoxyes
  • Result: (pass)
    • Test Case: Use down arrow (Read next item) to navigate forward to `*[role="alertdialog"]`
      1. Launch NVDA and Firefox.
      2. Navigate to the test page.
      3. Find the target element(s) that you will test against. Identify all elements that match this selector:*[role="alertdialog"]
        • If multiple elements match the selector, repeat this test for all instances. However, choose a single instance to report against. If you feel that the selector should be more specific, please open a GitHub Issue.
      4. Position and configure the screen reader so that the following conditions are met
        • Virtual focus is: before target
        • Keyboard focus is: before target
        • Mode is: auto (do not explicitly change the mode)
      5. Issue the command: down arrow (Read next item)
        • After issuing the command, virtual focus should be: target
      6. Record results for the relevant expectations
    • Output: target
Notes: An element in the dialog had to receive focus before the contents of the dialog would be read by NVDA.
OrcaFirefoxyes
  • Result: (pass)
    • Test Case: Use down arrow (Read next item) to navigate forward to `*[role="alertdialog"]`
      1. Launch Orca and Firefox.
      2. Navigate to the test page.
      3. Find the target element(s) that you will test against. Identify all elements that match this selector:*[role="alertdialog"]
        • If multiple elements match the selector, repeat this test for all instances. However, choose a single instance to report against. If you feel that the selector should be more specific, please open a GitHub Issue.
      4. Position and configure the screen reader so that the following conditions are met
        • Virtual focus is: before target
        • Keyboard focus is: before target
        • Mode is: auto (do not explicitly change the mode)
      5. Issue the command: down arrow (Read next item)
        • After issuing the command, virtual focus should be: target
      6. Record results for the relevant expectations
    • Output: "target"
TalkBackChromeyes
  • Result: (pass)
    • Test Case: Use Swipe right (or alt+right arrow) (Read next item) to navigate forward to `*[role="alertdialog"]`
      1. Launch TalkBack and Chrome.
      2. Navigate to the test page.
      3. Find the target element(s) that you will test against. Identify all elements that match this selector:*[role="alertdialog"]
        • If multiple elements match the selector, repeat this test for all instances. However, choose a single instance to report against. If you feel that the selector should be more specific, please open a GitHub Issue.
      4. Position and configure the screen reader so that the following conditions are met
        • Virtual focus is: before target
        • Keyboard focus is: before target
        • Mode is: auto (do not explicitly change the mode)
      5. Issue the command: Swipe right (or alt+right arrow) (Read next item)
        • After issuing the command, virtual focus should be: target
      6. Record results for the relevant expectations
    • Output: target
Voice Access (Android)Chromenot applicable-
Voice Control (iOS)Safarinot applicable-
Voice Control (MacOS)Safarinot applicable-
VoiceOver (iOS)Safariyes
  • Result: (pass)
    • Test Case: Use Swipe Right (Next item) to navigate forward to `*[role="alertdialog"]`
      1. Launch VoiceOver (iOS) and Safari.
      2. Navigate to the test page.
      3. Find the target element(s) that you will test against. Identify all elements that match this selector:*[role="alertdialog"]
        • If multiple elements match the selector, repeat this test for all instances. However, choose a single instance to report against. If you feel that the selector should be more specific, please open a GitHub Issue.
      4. Position and configure the screen reader so that the following conditions are met
        • Virtual focus is: before target
        • Keyboard focus is: before target
        • Mode is: auto (do not explicitly change the mode)
      5. Issue the command: Swipe Right (Next item)
        • After issuing the command, virtual focus should be: target
      6. Record results for the relevant expectations
    • Output: target
VoiceOver (macOS)Safariyes
  • Result: (pass)
    • Test Case: Use VO + Right arrow (Read next item) to navigate forward to `*[role="alertdialog"]`
      1. Launch VoiceOver (macOS) and Safari.
      2. Navigate to the test page.
      3. Find the target element(s) that you will test against. Identify all elements that match this selector:*[role="alertdialog"]
        • If multiple elements match the selector, repeat this test for all instances. However, choose a single instance to report against. If you feel that the selector should be more specific, please open a GitHub Issue.
      4. Position and configure the screen reader so that the following conditions are met
        • Virtual focus is: before target
        • Keyboard focus is: before target
        • Mode is: auto (do not explicitly change the mode)
      5. Issue the command: VO + Right arrow (Read next item)
        • After issuing the command, virtual focus should be: target
      6. Record results for the relevant expectations
    • Output: target
Windows Speech RecognitionChromenot applicable-

Extended Support for: alertdialog role: NOT automatically switch away from reading mode

These are less common combinations

ATBrowserHas SupportOutput
Dragon Naturally SpeakingIEnot applicable-
Dragon Naturally SpeakingFirefoxnot applicable-
JAWSEdgeunknown-
NarratorFirefoxunknown-
NarratorChromeunknown-
NarratorIEunknown-
NVDAIEunknown-
NVDAEdgeunknown-
TalkBackFirefoxunknown-
Voice Access (Android)Firefoxnot applicable-
VoiceOver (iOS)Chromeunknown-
VoiceOver (macOS)Chromeunknown-
VoiceOver (macOS)Firefoxunknown-
OrcaChromeunknown-
Voice Control (MacOS)Chromenot applicable-
Voice Control (MacOS)Firefoxnot applicable-
Windows Speech RecognitionEdgenot applicable-
Back to top

Dates and Versions of tested combinations

ATAT VersionBrowserBrowser versionOS versionDate tested
JAWS2018.1808.10Chrome7418092019-05-19
JAWS2018.1811.2IE11.25318092019-01-19
JAWS2018.1811.2Firefox6618092019-05-19
Narrator1809Edge44.1776318092018-01-18
NVDA2018.1808.10Chrome7418092019-05-19
NVDA2018.4.1Firefox64.0.218092019-01-19
Orca3.31.4Firefox69Ubuntu 19.042019-09-13
TalkBack7.2Chrome678.12018-07-21
VoiceOver (iOS)12.1.2Safari12.1.212.1.22018-01-18
VoiceOver (macOS)10.14.2Safari12.0.210.14.22018-10-26

History

  • 2019-01-18 Test created
  • 2019-09-13 Orca results and missing information added

This test is found in the following features:

Feedback and Contribute

We use our GitHub repository to manage our issue tracking. Please provide as much information as you can for issues, and please leave the id in the issue title intact.

Other

What are expectations and how are they determined?