Closed lseeman closed 7 years ago
Assigned to Jim Smith (@jim-work) https://www.w3.org/WAI/GL/wiki/SC_Managers_Phase1
"words or phrases that are most frequently used" I'm wondering if this isn't too subjective - can different authors and auditors (who may not even be familiar with the field/content, but simply be building/auditing this for somebody else) unambiguously determine this?
@jim-work Is there a PR ready to go for this?
Pull request https://github.com/w3c/wcag21/pull/105
The general recommendation to prefer the 'active' voice over 'passive' seems dubious to me. I think the appropriate choice of mode really depends on the desired expression in context. Passive expressions can be clearer and easier to understand than active ones. For more background from a linguist, I recommend "Fear and Loathing of the English Passive" by Geoffrey K. Pullum http://www.lel.ed.ac.uk/~gpullum/passive_loathing.html
The recommendation to use active voices isn't based on stylistic preferences of linguists, it is based on the needs of the communities listed under Benefits. The recommendation to use an active voice is based on the research cited, for example, Accessible Information Guidelines , where the guidelines have been derived empirically.
Hi Jim-work Thanks for the link. Glad to see there is no general advice not to use the 'passive voice' in the guidelines you linked to (or I have missed it). I'd still be interested in the kind of method used to back up these guidelines empirically. Is anything on the methodology used available online? Best, Detlev
Sent from phone
Am 10.02.2017 um 01:46 schrieb jim-work notifications@github.com:
The recommendation to use active voices isn't based on stylistic preferences of linguists, it is based on the needs of the communities listed under Benefits. The recommendation to use an active voice is based on the research cited, for example, Accessible Information Guidelines https://www.stroke.org.uk/sites/default/files/Accessible%20Information%20Guidelines.pdf%281%29.pdf , where the guidelines have been derived empirically.
— You are receiving this because you commented. Reply to this email directly, view it on GitHub, or mute the thread.
Hi Jim,
Thanks for that link, it is an interesting read.
I am struck by "Page 13", which says:
Check the reading level of your sentence
• Microsoft Word can check the readability of sentences
• Readability is a measure of how hard a sentence is to follow
• It is graded from 0, which is easy, to high such as 20, which is very difficult
• For people with aphasia we are aiming for a score of 5 or lower
• Use Flesch-Kincaid to check readability
• You need to add a full-stop at the end of your sentence
• This allows the software to read the sentence
• Go to page 24 for how to set up Flesch-Kincaid
• Use Flesch-Kincaid • Check the Grade Level
• Check you have a number of 5 or lower
• If the number is 6 or higher, you need to make your sentence simpler
However, earlier on this list, it was said that "Reading Level" is not sufficient. Comments were made that we should not be using that as a method of measurement. We now have conflicting guidance.
Which is it? Please help the non-experts understand.
I also think we will get serious resistance if we insist that all web content be written at a Grade 5 reading level. It is hard to achieve consistently. The language does not seem natural to average readers.
This email scores 5.3 in Flesch-Kincaid. It fails the requirement above.
If I asked all content writers to write like this, I think they would refuse.
Thanks.
JF
On Thu, Feb 9, 2017 at 6:46 PM, jim-work notifications@github.com wrote:
The recommendation to use active voices isn't based on stylistic preferences of linguists, it is based on the needs of the communities listed under Benefits. The recommendation to use an active voice is based on the research cited, for example, Accessible Information Guidelines http://url https://www.stroke.org.uk/sites/default/files/Accessible%20Information% 20Guidelines.pdf%281%29.pdf , where the guidelines have been derived empirically.
— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/w3c/wcag21/issues/42#issuecomment-278824303, or mute the thread https://github.com/notifications/unsubscribe-auth/ABK-cxtMNO2PjdSTyGw6F4k55A4361V2ks5ra7NwgaJpZM4K9IWn .
-- John Foliot Principal Accessibility Strategist Deque Systems Inc. john.foliot@deque.com
Advancing the mission of digital accessibility and inclusion
Reading level helps make it easier to read. This SC makes it easier to understand.The reading level algorithm is unhelpful for the examples given in the benefits section. Hor example we would prefer the phrase " hot or cold" rather then the term "mode" See blurb in the SC) reading level on useful at all in the scope given at the higher conformance levels such as interface components etc were the main issue is understandability. .
Debbie Dahl did exhaustive research on this. I am just giving a short highlight
All the best
Lisa Seeman
LinkedIn, Twitter
---- On Sat, 11 Feb 2017 17:06:07 +0200 John Foliot<notifications@github.com> wrote ----
Hi Jim,
Thanks for that link, it is an interesting read.
I am struck by "Page 13", which says:
Check the reading level of your sentence
• Microsoft Word can check the readability of sentences
• Readability is a measure of how hard a sentence is to follow
• It is graded from 0, which is easy, to high such as 20, which is very difficult
• For people with aphasia we are aiming for a score of 5 or lower
• Use Flesch-Kincaid to check readability
• You need to add a full-stop at the end of your sentence
• This allows the software to read the sentence
• Go to page 24 for how to set up Flesch-Kincaid
• Use Flesch-Kincaid • Check the Grade Level
• Check you have a number of 5 or lower
• If the number is 6 or higher, you need to make your sentence simpler
However, earlier on this list, it was said that "Reading Level" is not sufficient. Comments were made that we should not be using that as a method of measurement. We now have conflicting guidance.
Which is it? Please help the non-experts understand.
I also think we will get serious resistance if we insist that all web content be written at a Grade 5 reading level. It is hard to achieve consistently. The language does not seem natural to average readers.
This email scores 5.3 in Flesch-Kincaid. It fails the requirement above.
If I asked all content writers to write like this, I think they would refuse.
Thanks.
JF
On Thu, Feb 9, 2017 at 6:46 PM, jim-work <notifications@github.com> wrote:
> The recommendation to use active voices isn't based on stylistic > preferences of linguists, it is based on the needs of the communities > listed under Benefits. > The recommendation to use an active voice is based on the research cited, > for example, Accessible Information Guidelines <http://url> > https://www.stroke.org.uk/sites/default/files/Accessible%20Information% > 20Guidelines.pdf%281%29.pdf , where the guidelines have been derived > empirically. > > — > You are receiving this because you are subscribed to this thread. > Reply to this email directly, view it on GitHub > <https://github.com/w3c/wcag21/issues/42#issuecomment-278824303>, or mute > the thread > <https://github.com/notifications/unsubscribe-auth/ABK-cxtMNO2PjdSTyGw6F4k55A4361V2ks5ra7NwgaJpZM4K9IWn> > . >
-- John Foliot Principal Accessibility Strategist Deque Systems Inc. john.foliot@deque.com
Advancing the mission of digital accessibility and inclusion — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub, or mute the thread.
Detlev, Can I suggest you read the research module?
All the best
Lisa Seeman
Hi Lisa, I have read the research module https://www.w3.org/TR/coga-user-research/ While the text gives a good overview of a range of different cognitive disabilities and in turn usage issues including typical use cases, I find little on usability research that explores how successfully these usage issues can be alleviated by the web design recommendations given.
Many recommendations (good heading structure, short sentences, consistency, use of white space etc.) clearly make sense, others such as "Use active rather than passive voice" seem to me to be too general and therefore possibly counter-productive (just one example: 'Mrs Smith was taken to hospital' is simpler and clearer than 'An ambulance took Mrs Smith to hospital' - the first is passive, the second the recommended active voice).
I would also love to see empirical usability studies that investigate the use of plugins for adaptive interfaces that can hide and reveal parts of the interface. As others have noted these may reduce complexity once understood but they also add complexity. My specific request was whether there are empirical usability studies with users with learning and cognitive disabilities (preferably online) that support the advice given to web authors.
Best, Detlev
Sent from phone
Am 12.02.2017 um 12:44 schrieb Lisa Seeman notifications@github.com:
Detlev, Can I suggest you read the research module?
All the best
Lisa Seeman
— You are receiving this because you commented. Reply to this email directly, view it on GitHub, or mute the thread.
This proposal is difficult to measure and to implement. I recommend looking at using reading level. It isn't perfect, but it addresses most of the user needs identified, especially when paired with existing Technique G153. Reading level has international support, it has automated tests, and it has a variety of formulas (Flesh-Kincaid is the oldest and best known, there are many others.) The existing WCAG AAA 3.1.5 could be better adapted (and revised for clarity) to apply to all content.
I changed the handle to provide a more clear progression with the other suggestions for revision to the Plain Language success criteria proposals: Issue 30 - Understandable Labels, and Issue 41: Understandable Instructions.
Proposed revision: Understandable Content: Blocks of text either: (AAA)
[links to WCAG’s definitions of lower secondary education and blocks of text]
Current versions of SC and Definitions
SC in full draft guideline Note these terms are also used for #41 so be sure not to wind up with colliding definitions.
Plain language (All Content)
Plain language: Allow the user to use plain language and provide clear and simple language in all content so that all of the following are true:
Exceptions:
What Principle and Guideline the SC falls within.
Under WCAG 3.1
Suggestion for Priority Level
AAA or AA depending on whether technology can ensure that this is reasonable via alternative content.
Related Glossary additions or changes
Description
The intent of this Success Criterion is to ensure people can understand and use all content. For language to be usable by many people with cognitive disabilities it needs to be written in plain, clear language. This is an important accessibility principle.
It should be noted that the exceptions ensure it is widely applicable.
Benefits
This success criterion supports those who have reading difficulties, language disabilities, and some visual perceptual difficulties. It can include people with intellectual disabilities, receptive aphasia, and/or acquired dyslexia, as well as those with general cognitive learning disabilities. This supports those who have dementia, and/or acquire cognitive disabilities as they age.
Related Resources
Stroke Association Accessible Information Guidelines http://www.stroke.org.uk/professionals/accessible-information-guidelines
Computers helping people with special needs, 14 international conference ICCHP 2014 Eds. Miesenberger, Fels, Archambault, et al. Springer (pages 401). Paper: Never Too Old to Use a Tablet, L. Muskens, et al. pages 392 - 393.
[i.49] Vogindroukas, I. & Zikopoulou, O. (2011). Idiom understanding in people with Asperger syndrome/high functioning autism. Rev. soc. bras. fonoaudiol. Vol.16, n.4, pp.390-395.
NOTE: Available at http://www.scielo.br/scielo.php?script=sci_arttext&pid=S1516-80342011000400005&lng=en&nrm=iso .
[i.50] Oi, M., Tanaka, S. & Ohoka, H. (2013). The Relationship between Comprehension of Figurative Language by Japanese Children with High Functioning Autism Spectrum Disorders and College Freshmen's Assessment of Its Conventionality of Usage, Autism Research and Treatment, vol. 2013, Article ID 480635, 7 pages, 2013. doi:10.1155/2013/480635.
NOTE: Available at http://www.hindawi.com/journals/aurt/2013/480635 /.
[i.51] de Villiers, P. A. et al. (2011). Non-Literal Language and Theory of Mind in Autism Spectrum Disorders. Poster presented at the ASHA Convention, San Diego.
NOTE: Available at http://www.asha.org/Events/convention/handouts/2011/de-Villiers-de-Villiers-Diaz-Cheung-Alig-Raditz-Paul/ .
[i.52] Norbury, C. F. (2005). The relationship between theory of mind and metaphor: Evidence from children with language impairment and autistic spectrum disorder.; Oxford Study of Children's Communication Impairments, University of Oxford, UK; British Journal of Developmental Psychology, 23, 383-39.
NOTE: Available at http://www.pc.rhul.ac.uk/sites/lilac/new_site/wp-content/uploads/2010/04/metaphor.pdf.
[i.53] Language and Understanding Minds: Connections in Autism; Helen Tager-Flusberg, Ph.D; Chapter for: S. Baron-Cohen, H. Tager-Flusberg, & D. J. Cohen (Eds.), Understanding other minds: Perspectives from autism and developmental cognitive neuroscience. Second Edition. Oxford: Oxford University Press.
NOTE: Available at http://www.ucd.ie/artspgs/langimp/TAG2.pdf.
Phiriyapkanon. Is big button interface enough for elderly users, P34, Malardardalen University Press Sweden 2011.
Neilson-aging
Top Five Instructional Tips for Students with Down Syndrome"http://specialedpost.org/2013/01/31/top-five-instructional-strategies-for-students-with-down-syndrome/
http://www.autism.org.uk/working-with/autism-friendly-places/designing-websites-suitable-for-people-with-autism-spectrum-disorders.aspx (downloaded 08/2015)
Students with Down Syndrome, http://www.downssa.asn.au/__files/f/3203/A%20Student%20with%20Down%20Syndrome%202014.pdf
Task force links
Issue papers
Testability
The success criterion is testable if each of the bullet points are testable. If all the content fails any bullet point, it is not conformant to this success criterion. If it passes all of the bullet points, it is conformant.
Bullet points:
Tense and voice are objective, and hence are verifiable. Also, it is expected that natural language processing algorithms will be able to confirm this automatically with reasonable accuracy.
Testing for exceptions:
If present tense and active voice have not been used, the tester will need to confirm if one of the exceptions is relevant. If an exception is not relevant, and present tense and active voice have not been used, then the content fails this success criterion.
Even languages with a small number of users have published lists of the most frequent words (such as Hebrew). If there is a natural language that does not have one, algorithms exist that calculate these lists for a language, or for specific contexts. Testing content against these word lists can be done manually. However, it is expected there will be a natural language processing testing tool by the time this goes to CR. (It is already integrated into a tool by IBM.)
Testing for exceptions is as discussed above.
Use of double negatives is a fact, and hence is verifiable. It is assumed a natural language processing tool will also test for this. Testing for exceptions is as discussed above.
Non-literal text and metaphors can be identified when the meaning of the sentence is something other than the meaning of the individual words. This is human testable. Cognitive computing algorithms can test for this as well.
If the text is not literal, then the tester must confirm that personalization and an easy user setting enables it to be replaced, such that all meaning is retained.
Techniques