E-mail List Archives

Search E-mail List Archives for

Number of posts in this thread: 6 (In chronological order)

From: Glarst
Date: Mon, Oct 21 2024 8:32AM
Subject: SCOPE
No previous message | Next message →

I've been asked to provide a real-world example of a barrier caused by the
absence of SCOPE on an otherwise conformant TH tag. This pertains
specifically to the recent Title II Final Rule, so I guess the bigger
question is whether SCOPE is required for conformance.
So before I go on a wild goose chase evaluating every table on the WWW,
does anyone have a hot take on this (or a real-world instance they'd like
to share)?
Thanks,
Glarst

From: Hayman, Douglass
Date: Mon, Oct 21 2024 8:56AM
Subject: Re: - SCOPE
← Previous message | Next message →

I'd say, make two tables, one with scope, one without for both a web page and something like a PDF file and then navigate through the table with NVDA or JAWS to see the real world experience.

Doug Hayman
IT Accessibility Coordinator
Information Technology
Olympic College
= EMAIL ADDRESS REMOVED =
(360) 475-7632

From: Laura Roberts
Date: Mon, Oct 21 2024 9:25AM
Subject: Re: SCOPE
← Previous message | Next message →

Use a screen reader on a table with a ton of data cells.

Talk about how if we lose our place, we can see where we are at a glance by
looking at where the row header and column header and data cell intersect.
Ask how could you tell where you are in this table without being able to
see? (Answer is you have to listen to the whole damn thing again)

Then use the screen reader say cell contents command on a table with and
without scope set. (CTRL ALT Numpad 5 in Jaws). Becomes crystal clear why
it's needed.

On Mon, Oct 21, 2024, 10:33 AM Glarst < = EMAIL ADDRESS REMOVED = > wrote:

> I've been asked to provide a real-world example of a barrier caused by the
> absence of SCOPE on an otherwise conformant TH tag. This pertains
> specifically to the recent Title II Final Rule, so I guess the bigger
> question is whether SCOPE is required for conformance.
> So before I go on a wild goose chase evaluating every table on the WWW,
> does anyone have a hot take on this (or a real-world instance they'd like
> to share)?
> Thanks,
> Glarst
> > > > >

From: Jared Smith
Date: Mon, Oct 21 2024 9:31AM
Subject: Re: SCOPE
← Previous message | Next message →

At WebAIM we recently conducted testing on a wide variety of tables to answer this precise question. For basic tables most browser and screen readers properly assume the correct header scope. However, in cases where the top left cell is a row header, where a table has only row headers (no column headers), or when spanned header cells are present, implementation of the scope attribute provides a much more accessible experience in several cases. A more notable finding is that table accessibility support is still quite poor, even when scope is properly implemented.

The test cases are at https://webaim.org/temp/tabletest.htm

The test results for VoiceOver, VoiceOver Mobile, and NVDA are at
https://docs.google.com/spreadsheets/d/1QNjleey20PzIB9s814uPQHQFzq9c25UkPJl7K_rzvVQ/edit?usp=sharing The highlighted cells are those in which scope is properly implemented, but the support is not good.

In short, we still recommend the scope attribute because it has positive accessibility benefits, though in some cases tables are accessible regardless, and in some other cases the tables are inaccessible regardless.

Jared Smith

From: Dax Castro
Date: Mon, Oct 21 2024 9:32AM
Subject: Re: SCOPE
← Previous message | Next message →

Any table that has row headers would provide a poor user experience. When a person is moving from row to row. Because without the scope of row for the items in the first column, the information would not be voiced correctly. This is a clear violation of 1.3.1 info and relationships.

Also, any table that has more than one column header would not voice correctly without scope as well. By default, NVDA and JAWS only assume that the first row contain column headers. Therefore, any table that exceeds this simple design will fail for the end user to provide an accurate experience.

Thanks,
Dax
From: WebAIM-Forum < = EMAIL ADDRESS REMOVED = > on behalf of Laura Roberts < = EMAIL ADDRESS REMOVED = >
Sent: Monday, October 21, 2024 8:25:30 AM
To: WebAIM Discussion List < = EMAIL ADDRESS REMOVED = >
Subject: Re: [WebAIM] SCOPE

Use a screen reader on a table with a ton of data cells. Talk about how if we lose our place, we can see where we are at a glance by looking at where the row header and column header and data cell intersect. Ask how could you tell where you
NkdkJdXPPEBannerStart
Be Careful With This Message
From (Laura Roberts < = EMAIL ADDRESS REMOVED = >)<https://godaddy1.cloud-protect.net/email-details/?k=k1&payload=53616c7465645f5fa267a40fcb1fe78fbd93fd9e086ec2c19d07abdab8582f893f957c13d885eda65d43ddc55150f3ed38a607994f564ac0110a6f88aab7269d183f3017cf20eb352626fd8d5ca1a82527d1b0b36cb234f5c046e28aee0c26d85b717eb7fe5727cb23a764d2fe164a5c7aabff48c42b2adf25cbba52d15f50fad7318bb9a37ea96d6e5ba769f456f9eee83a10803e981ceb114c06941dd1e69648de6664abf83231c4817aa562dfd8aa880254c8054f37c03568900c7b1f650dfe116508bf13dc37608ffa311c22782eae12ce844c5337d199d905f7d4bbbc67>
Learn More<https://godaddy1.cloud-protect.net/email-details/?k=k1&payload=53616c7465645f5fa267a40fcb1fe78fbd93fd9e086ec2c19d07abdab8582f893f957c13d885eda65d43ddc55150f3ed38a607994f564ac0110a6f88aab7269d183f3017cf20eb352626fd8d5ca1a82527d1b0b36cb234f5c046e28aee0c26d85b717eb7fe5727cb23a764d2fe164a5c7aabff48c42b2adf25cbba52d15f50fad7318bb9a37ea96d6e5ba769f456f9eee83a10803e981ceb114c06941dd1e69648de6664abf83231c4817aa562dfd8aa880254c8054f37c03568900c7b1f650dfe116508bf13dc37608ffa311c22782eae12ce844c5337d199d905f7d4bbbc67>
Potential Impersonation
The sender's identity could not be verified and someone may be impersonating the sender. Take caution when interacting with this message.
NkdkJdXPPEBannerEnd

Use a screen reader on a table with a ton of data cells.

Talk about how if we lose our place, we can see where we are at a glance by
looking at where the row header and column header and data cell intersect.
Ask how could you tell where you are in this table without being able to
see? (Answer is you have to listen to the whole damn thing again)

Then use the screen reader say cell contents command on a table with and
without scope set. (CTRL ALT Numpad 5 in Jaws). Becomes crystal clear why
it's needed.

On Mon, Oct 21, 2024, 10:33 AM Glarst < = EMAIL ADDRESS REMOVED = > wrote:

> I've been asked to provide a real-world example of a barrier caused by the
> absence of SCOPE on an otherwise conformant TH tag. This pertains
> specifically to the recent Title II Final Rule, so I guess the bigger
> question is whether SCOPE is required for conformance.
> So before I go on a wild goose chase evaluating every table on the WWW,
> does anyone have a hot take on this (or a real-world instance they'd like
> to share)?
> Thanks,
> Glarst
> > > > >

From: Birkir R. Gunnarsson
Date: Tue, Oct 22 2024 3:55AM
Subject: Re: SCOPE
← Previous message | No next message

Any <th> cell in the first row in a table has an implicit scope of "col".
(i.e. column header)
Any <th> cell in the first column in a table (not counting the first row)
has an implicit scope of "row".
Adding a scope attribute to those cells typically does not change
accessibility support, in my experience.
(I have not yet looked at the latest experiments from the mighty WebAIM
team)


You do need a scope attribute for
* Any <th> cell that is not in the first row or first column in the table,
(scope="col" for a column header, scope="row" for a row header)
* Any <th> cell that spans multiple columns or rows,
* use the scope attribute with values of colgroup or rowgroup respectively
and use the colspan or rowspan attributes to indicate number of columns or
rows spanned
* <th scope="colgroup" colspan="3">I am a column header cell spanning 3
columns</th>
* <th scope="rowgroup" rowspan="3">I am a header cell spanning 3 rows</th>
You need to do this regardless of the location of the <th> cell in the
table.

I disagree that a table with only column or row headers fails WCAG 1.3.1,
it all depends on
I) Whether a cell is visually emphasized
II) If the table has a natural candidate

A table needs either column or row header cells, usually it needs column
header cells, but it does not necessarily need both.

Since screen readers read header cells in context, you don't want those
cells to be too wordy.

Imagine that you have a table of transactions. The first column is a
30-digit transaction ID, Then you have the columns date, amount, from
account and to account.
The transaction number doesn't tell you anything, and you don't want to be
forced to hear it every time you navigate between rows, so it's a terrible
candidate for a row header.
Any one or more of the amount, date, or from account columns would be
acceptable row header candidates, though they don't tell the whole story,
probably date and amount would work together as row headers

I would personally prefer not having a row header in that table.

Actually what I do want is for screen reader users to set their own row
header cells, but that would be a screen reader functionality.