W3C

Results of Questionnaire EOWG Weekly Survey 7 January 2015

The results of this questionnaire are available to anybody. In addition, answers are sent to the following email addresses: shawn@w3.org, ee@w3.org

This questionnaire was open from 2015-01-07 to 2015-01-09.

8 answers have been received.

Jump to results for question:

  1. Tables Tutorial: th replacement
  2. Tables Tutorial: Scope attribute on all but most basic tables.

1. Tables Tutorial: th replacement

Summary:
A reviewer suggests that we change information in the tutorial to account for <td scope="row|column"> being used instead of <th> in HTML4.
However, some say this is not best practice. Also, it's not allowed in HTML5.

Location: Multi-directional Tables → Table with an offset column of header cells

Background: The reviewer writes:

If the doctype is HTML4, td with scope is valid and breaks nothing. Please refer to the reasoning in the HTML4 specs . In fact TH in the second column may not be agreeable to some UI designers (even with HTML5) because they may not want centered and bolded content in the middle of a table. Using CSS to do away the visual effect of TH negates the proper use of TH. So when it does not suit UI design, using TD and scope may be the way to go ... sure it may not be valid in HTML5 but that does not fail SC 4.1.1 or any WCAG2 SC.

Further reading: E-mail 6 Jan

Related information:

  • F91: Failure of Success Criterion 1.3.1 for not correctly marking up table headers says: (emphasis mine)

    This failure occurs when data tables do not use header elements (th) or other appropriate table mark-up (the scope attribute, headers and id or the ARIA roles columnheader / rowheader) to make the headers programatically determinable from within table content.

    The failiure lists td with scope attribute as the 5th of six tests to find out if table markup is correct.

  • H63: Using the scope attribute to associate header cells and data cells in data tables says:

    Based on screen reader support today, [scope] use is suggested in two situations both relating to simple tables: […] header cells marked up with td instead of th. Sometimes, authors use this to avoid the display characteristics associated with th and also do not choose to use CSS to control the display for th.

EOWG input:
Please put below your thougths on whether or not we should change the text to allow for <td scope="row|column"> being used instead of <th> in certain circumstances like avoiding <th> styles.

Details

Responder Tables Tutorial: th replacement
Shawn Lawton Henry *If* I'm understanding the issue fully: No, we should not. 1. I think we should suggest best practice, knowing about HTML5. If something is not allowed in HTML5 and not needed in HTML4, then I don't think we should advice it at all. 2. I do not agree that "Using CSS to do away the visual effect of TH negates the proper use of TH." I can change the visual styling of TH and still make it visually clearly a heading. (and, in fact, I think undoing centering is usually important for usability and maybe for accessibility for some folks as well)
Anna Belle Leiserson I think it's better to leave the text as is -- for the same reasons Shawn has given.
Sharron Rush Recommend no change to text
Vivienne Conway
Paul Schantz I agree with Shawn. In the context of a tutorial, I think it's best to use examples that will work with both HTML4 and HTML5.
Helle Bjarnø I've been offline for so long, have to get back into EO work. Abstain
Kevin White The tutorials seek to promote HTML5 so I don't feel that adopting code that does not meet that spec would be appropriate. I would not support using SCOPE with TD.
Wayne Dick <th> should be used in headers.
Scope is required.

2. Tables Tutorial: Scope attribute on all but most basic tables.

Summary:

In the tutorials we recommend the use of the scope attribute when defining headers as we have found out that some screen readers interpret th elements as header cells for other ths in the same row/column.
The reviewer says:

6.

"If the table is larger or its content is more ambiguous, the scope attribute should be used to avoid confusion".
And in example 1: "The following table of concerts has the cells in the first row marked up as <th> cells without any scope direction. This is only acceptable because it is such a small table and the data itself is distinctly different in each column.Note: Some screen readers will read “Date – Event – Venue” on the “Venue” cell because the direction of the <th> elements is ambiguous".

Comment: "larger or its content is more ambiguous" is itself vague/not well defined.
Use of scope is not warranted merely based on number of rows or columns. Secondly, for a simple table with TH in first row and/or first column, the scope attribute really does not add value and is unnecessary. It is merely redundant markup. (It only prevented older versions of JAWS from reading all TH cells above current cell in the first column of TBODY as header cells ).
Also older versions of JAWS(before JFW 15) may have read all cells to left of current cell in the first row as headers. This is a bug and is not the case now. I had pointed this out to FS a while ago.
Please refer to definition or description of "data table" in comment#1 above. Every column of a data table indeed presents a distinct attribute so it is not useful to justify not using scope with "such a small table and the data itself is distinctly different in each column".
The only instance where scope is needed in a simple data table is where the row identifier is not in the first column like in "Example 4: Table with an offset column of header cells".

Further reading: Dec 31st email

Location: Tables Tutorial → Table with header cells in the top row only

Background: The topic was discussed in several meetings, including:

EOWG input:
Please put below your thoughts on whether or not we should change examples with only one header row/column to not include scope.

Details

Responder Tables Tutorial: Scope attribute on all but most basic tables.
Shawn Lawton Henry I think probably leave scope. While we say that you shouldn't make accessibility *requirements* to account for AT bugs, I think the tutorials are different -- they are to be *best practice*, and sometimes that means accounting for bugs.
Reviewer says: "older versions of JAWS(before JFW 15)". That's not very old. We know people have older versions of screen readers. (and some of us have O/Ss and browsers wayy older than that :-)
Anna Belle Leiserson I don't know much about screen readers, so defer.
Sharron Rush Reviewing Bim's comment, she said "I believe that scope is generally needed to assure clarity for screen readers" but that it was deemed optional and not clearly a requirement in H63. Recommend to leave the examples as is but make the clarification in the text that it is optional.
Vivienne Conway Sorry, difficult for me to say as I wasn't part of the original discussion. However from my reading of the discussions, I think the EOWG input on not including the scope in the simple example one would be okay.
Paul Schantz Like Annabelle, I'm not knowledgeable about screen readers, so I defer this to those who are.
Helle Bjarnø I've been offline for so long, have to get back into EO work. Abstain
Kevin White Given that there is little harm (slightly more code) in including SCOPE and the potential to improve user experience if it is included, I can't see why we would not include it. Perhaps it warrants a little bit of investigation as to what browser/screenreader/OSs it is of benefit for... although this is quite a bit of work.
Wayne Dick I would always include scope. There is always the first cell ambiguity.

More details on responses

  • Shawn Lawton Henry: last responded on 8, January 2015 at 19:17 (UTC)
  • Anna Belle Leiserson: last responded on 8, January 2015 at 21:43 (UTC)
  • Sharron Rush: last responded on 8, January 2015 at 23:13 (UTC)
  • Vivienne Conway: last responded on 9, January 2015 at 02:22 (UTC)
  • Paul Schantz: last responded on 9, January 2015 at 05:37 (UTC)
  • Helle Bjarnø: last responded on 9, January 2015 at 07:52 (UTC)
  • Kevin White: last responded on 9, January 2015 at 10:23 (UTC)
  • Wayne Dick: last responded on 9, January 2015 at 13:44 (UTC)

Non-responders

The following persons have not answered the questionnaire:

  1. Eric Velleman
  2. Andrew Arch
  3. Shadi Abou-Zahra
  4. Sylvie Duchateau
  5. Kazuhito Kidachi
  6. Jedi Lin
  7. David Sloan
  8. Mary Jo Mueller
  9. Vicki Menezes Miller
  10. Reinaldo Ferraz
  11. Bill Kasdorf
  12. Cristina Mussinelli
  13. Kevin Rydberg
  14. Adina Halter
  15. Denis Boudreau
  16. Laura Keen
  17. Sarah Pulis
  18. Bill Tyler
  19. Gregorio Pellegrino
  20. Ruoxi Ran
  21. Jennifer Chadwick
  22. Sean Kelly
  23. Muhammad Saleem
  24. Sarah Lewthwaite
  25. Daniel Montalvo
  26. Mark Palmer
  27. Jade Matos Carew
  28. Sonsoles López Pernas
  29. Greta Krafsig
  30. Jason McKee
  31. Jayne Schurick
  32. Billie Johnston
  33. Michele Williams
  34. Shikha Nikhil Dwivedi
  35. Brian Elton
  36. Julianna Rowsell
  37. Tabitha Mahoney
  38. Fred Edora
  39. Rabab Gomaa
  40. Marcelo Paiva
  41. Eloisa Guerrero
  42. Leonard Beasley
  43. Frankie Wolf
  44. Supriya Makude
  45. Aleksandar Cindrikj
  46. Angela Young

Send an email to all the non-responders.


Compact view of the results / list of email addresses of the responders

WBS home / Questionnaires / WG questionnaires / Answer this questionnaire