RE: Change in Context?

This does not sound like a change of context to me. This is defined at https://www.w3.org/TR/UNDERSTANDING-WCAG20/consistent-behavior-unpredictable-change.html#context-changedef

There is clearly no change of user agent, viewport or focus. It also sounds like there is a change of content, but not a change of "content that changes the meaning of the Web page".

As long as you can get to all the content using keyboard navigation, then it doesn't sound like there are any WCAG non-conformances. That said, it sounds like a pretty ugly user experience. Did that design evolve from user research? I'm guessing not. Or did some designer dream it up out of nowhere? Now that sounds more likely.

It would be interesting to do some quick and dirty user testing with screen readers, Dragon users and people who use keyboard navigation. That's the real measure of how accessible it is.

Steve Green
Managing Director
Test Partners Ltd


From: andrew.nordlund@hrsdc-rhdcc.gc.ca <andrew.nordlund@hrsdc-rhdcc.gc.ca>
Sent: 16 September 2020 12:19
To: w3c-wai-ig@w3.org
Subject: Change in Context?

Hello,

I'm looking at an interesting scenario that's caused a bit of an internal debate, and wondering if it's constitutes an accessibility problem (ie: a WCAG failure):

When you load the page in question, all you see is the main content starting with an <h1> and then a paragraph, etc.  So far so good.

What happens next depends on how you navigate:
A) If you scroll down with your mouse, arrow keys, or page-down key then after a short amount of scrolling, a header section appears at the top of the page.  (The header section includes the name and icon of the organization, a few links including "About", "Contact Us", etc. which don't appear elsewhere in the page.)
B) But if you start navigating by hitting the Tab key then the first things you tab to are skip links ("Skip to main content", "Skip to FAQ", etc.) which become visible as you tab to them (and disappear as you tab away).  Then you tab to the header section described in A.  As you hit Tab, you tab to all the links in the header section.  Then you tab to the first tabstop in the main content, and the header section remains visible.

If, upon loading the page, you click in the main content, then hit Shift+Tab then the header section becomes visible, and your focus goes to the last link in the header.

Checking with a screen reader (NVDA+Firefox/Chrome):
Upon loading the page, if you start reading the page top-to-bottom, or using the down arrow key you hear the skip links then the <h1> then the first paragraph, etc.  No mention of the header section.  But if you start navigating by using the Tab key, then you hear the skip links, then the links in the header section.

If you bring up a list of links on the page (NVDA key + F7) if the header section is visible, then its links will be listed..  If the header is not visible, then the header links will not be in the list.


There's something about this that doesn't feel right, but I'm not sure if it's a WCAG violation.  Is scrolling causing a change in context?  Is this an accessibility problem?  Is this a WCAG failure?  If so, what Success Criterion does this miss?


Notes:
* I'm using the words "header section" loosely.  The section is not in a <header> element nor does it have role='banner'.  It's just at the top in a <div> with no roles, or specific semantic markup.
* At all times the focus indicator is visible.  This is not a 2.4.7 - Focus Visible violation.


-=Andrew Nordlund