– DRAFT –
Cognitive and Learning Disabilities Accessibility Task Force Teleconference

<Lisa> zakim pick a scribe

<Lisa> https://www.w3.org/WAI/GL/task-forces/coga/wiki/Scribe_list

<Lisa> next item

<Lisa> close item 1

<Lisa> next item

<julierawe> Rain Structure will meet this Thursday and go over recruiting protocol

<julierawe> Rain I will have draft ready to share the day before the meeting

<julierawe> Rain This is to get ready to start user study

<julierawe> Rain I haven't heard from Roy about our requests for new version

<julierawe> Rain My colleague is working on more idealized version, so those two prototypes are in development now

<Lisa> https://www.w3.org/WAI/GL/task-forces/coga/wiki/Developer_resources#Resources_from_other_projects

<julierawe> Lisa We have a wiki page on resources from other projects including Easy Reading project

<julierawe> Lisa We had a nice handbook for involving peer researchers and focus groups

<julierawe> Lisa This includes an example of a consent form

<julierawe> Lisa Update on the issue papers—Becca is doing online safety and Julie is doing the two others

<julierawe> Becca_Monteleone I just plugged in a google doc version and that's where I'm doing the edits. I put the link in the Action Items doc.

<julierawe> Lisa We're doing something similar

<julierawe> julierawe I will create google docs and edit them this week. I will also share with Becca the COGA style guide that includes details on how many words per sentence.

<Becca_Monteleone> Rain - I also have experience with cognitively accessible consent/assent forms and navigating ethics around research, particularly with people with intellectual disabilities in an academic setting, if that's helpful!

<Rain> +1 it surprised me as well

<Jennie_Delisi> I have been adding the labels subgroup during the AG calls.

<julierawe> julierawe For Clear Language, Julie will share Rain's card-sort activity. Implied Language will continue meeting during the Tuesday AG meetings.

<Jennie_Delisi> Sorry - been attending the labels subgroup

<julierawe> JMcSorley I did not know the sprints had started

<julierawe> Rain I got lucky and realized they started early

<julierawe> Lisa Julie, can you ask COGA if everyone was contacted by the subgroup they wanted to be in

<julierawe> Rain I want to acknowledge there have been a couple of challenges for the COGA task force in terms of participating.

<julierawe> Rain I'd also like to acknowledge they are trying to find a way to be more effective. Giving them feedback will be valuable but I want to recognize they are trying.

<Jennie_Delisi> +1 to Rain's comment. Because they made the change, I have more opportunity to participate. But do support providing the feedback

<julierawe> Lisa We need to catch these things when people are getting stuck

<julierawe> JMcSorley I get a lot of emails from W3C so it could be an error on my side as well

<julierawe> julierawe I think this is partly a failure on my part—I should have communicated more clearly

<julierawe> tburtin I think they were also trying to respond to scheduling challenges and decided to use AG time

<Lisa> next item

<julierawe> julierawe I will reach out to AG to see if subgroup leaders can get a list of survey takers who expressed interest

<Jennie_Delisi> September 2 is holiday in the US and Canada

<Lisa> last week august and 1st week septber

<Rain> +1

<julierawe> +1

<Becca_Monteleone> +1

<Jennie_Delisi> +1

<tburtin> +1

<JMcSorley> +1

<kirkwood> +1

<Lisa> next item

<julierawe> No COGA meetings that last week of August and 1st week of September

<julierawe> Lisa CSUN is coming up in March

<Lisa> https://w2.csun.edu/cod/conference/general-call-presentations

<julierawe> Lisa The call for CSUN submissions is Sept 24. That might be another time for face-to-face meeting

<Lisa> next item

<Jennie_Delisi> TPAC starts September 23

<Lisa> https://www.w3.org/news-events/w3c-tpac/

<Lisa> next item

<Lisa> https://w3c.github.io/ctaur/

<julierawe> Lisa There is a new editor's draft and they want to know by end of September if we have any outstanding issues.

<kirkwood> +1 to a working or regular call

<julierawe> Lisa They think they have resolved our concerns. Should we set up a meeting or read it over if we think there are issues?

<kirkwood> +1

<julierawe> Lisa We should do it altogether

<julierawe> Lisa David is the liaison, and Jennie, Lisa, and John K did a lot of work on it.

<julierawe> Jennie_Delisi I think David is closest to the latest version of it.

<julierawe> Lisa I think we should loop in David but also set up a working meeting.

<julierawe> +1

<Lisa> next item

<julierawe> Lisa I sent an email about booking a room. I need to find out if we have quorum.

<julierawe> Frankie and ??? are definitely going

<julierawe> Lisa Frankie and Rain are going

<julierawe> Lisa Maybe from Julie, John K, Len and Tiffany

<julierawe> tburtin We are trying to get funding

<julierawe> Lisa Quorum is meant to be 6 people. 5 is OK. In a pinch, 4 might be OK with 3 remote people.

<julierawe> Lisa I will ask John Rochford

<julierawe> I should know in a week or so if I can attend

<julierawe> Lisa For now we have not booked a space

<julierawe> Lisa But that does not mean it's not worth going because we can still do breakout sessions and attend AG and APA meetings

<julierawe> Lisa If two more people manage to get space, we can ask for a room

<tburtin> How often does TPAC occur in the US?

<Lisa> https://www.w3.org/2024/09/TPAC/breakouts.html

<julierawe> Lisa I don't feel comfortable at this point booking a space and then other groups can't get that space

<julierawe> tburtin How often does TPAC come to the US?

<julierawe> Lisa I think it's every other year

<julierawe> Lisa But it might be every 3rd year

<julierawe> Lisa Rain, you wanted to do a breakout on the new structure for Making Content Usable

<julierawe> Rain Yes, we'll have some results from the qualitative testing and an opportunity to connect directly with the folks at W3C who control the formatting

<julierawe> Rain We'll be able to figure out what's possible as far as W3C style guide goes and how to address results from the testing

<julierawe> Lisa We're thinking about 4 breakouts: structure proposal, safety proposal and algorithms, supported decision-making, and triggers.

<julierawe> Lisa Deadline is Sept 15

<julierawe> Lisa Does anyone want to co-lead with me?

<julierawe> kirkwood Yes, supported decision-making

<julierawe> Lisa Great, let's try to get that paper ready in time

<julierawe> Lisa We'll have a presence at TPAC, we'll have breakout sessions, we might meet but it's a bit of a long shot

<Lisa> next item

<Lisa> review supported dessions making . See https://docs.google.com/document/d/1g9N-_3l-d6t0ppZcIIZ98k00x7ZE7h3ayq0OCX9Ycu4/edit#heading=h.rhs6n8lpznbz [from lisa]

<Lisa> https://docs.google.com/document/d/1g9N-_3l-d6t0ppZcIIZ98k00x7ZE7h3ayq0OCX9Ycu4/edit#heading=h.gvof99kcf7wa

<julierawe> Lisa Frankie and I will try to answer the questions this week

<julierawe> Lisa Normally we put in editor's draft when they're a bit more mature.

<julierawe> Lisa We could give people a week or two for people to weigh in before we put in editor's draft. What do people think?

<julierawe> Jennie_Delisi This morning I put into the list that the state of Minnesota is putting together a task force and they specifically mention guardianships and conservatorships

<julierawe> Jennie_Delisi This means in the future there will be groups to share this information with

<julierawe> Lisa This means getting the editors draft soon is all the more urgent

<julierawe> Lisa Let's not wait a year to share with them

<julierawe> Jennie_Delisi Their first meeting is in a year

<julierawe> kirkwood I do think the supported decision making needs a little more work

<julierawe> Lisa Can we get it ready in 2 weeks?

<julierawe> kirkwood I will set aside time to work with Jennie and others

<julierawe> Lisa I will set up a working call

<Jennie_Delisi> Apologies - I cannot add any meetings in the next 2 weeks

<julierawe> kirkwood We're not really describing the principles of cognitive support

<julierawe> kirkwood It needs to move further and also have Jennie review

<julierawe> Jennie_Delisi It would be easier to do asynchronous review

<julierawe> Jennie_Delisi If there's something you're hoping I look at in a couple days, just shoot me an email

<julierawe> Jennie_Delisi I do not see tagged comments during the work day

<Lisa> next item

<Lisa> close item 8

<julierawe> The Abstract explains that:

<julierawe> "This document is for people who make web content (web pages) and web applications. It gives advice on how to make content usable . . . "

<julierawe> but does not explain and rarely even mentions the role of user agents (browsers + assistive technology). If it is expected that authors (designers and developers of content and application) best meet all these "8 User Story objectives" and all the "8 Design Objectives" without relying on browser extenions or ATs (user agents), then that needs to be

<julierawe> made explicit.

<julierawe> If however, and this is recommended, some spectrum or trade-off beetween authors and user agents (UA) is a good, better, or best approach in meeting (or partially meeting) the specific objective, then that needs to be made explicit. I would recommend leaning towards the latter and being explicit when trying to meet the specific objective and

<julierawe> include considerations for the UA to be included. Perhaps it will be determined that authors have more responsibility of certains types of content and applications and UA's have more responsibilities on other types of content and applications. Examples would be helpful.

<julierawe> I need to understand the language used, including vocabulary, syntax, tense, and other aspects of language.

<julierawe> For example, the author would be responsibe for determing the initial choices of terms and levels of vocabulalry, but the UA would provide the hooks and mechanism to get to definitions or to adaptations of the terms (whole chucks of content) and alternative levels of the vocabulary (or whole chucks of content). Such as a simplified version of

<julierawe> "Terms and conditions" aternative vs the legally binding versions of "terms and conditions".

<julierawe> JMcSorley I think he is expecting that assistive technologies are going to be as robust as some of the screenreader technology is for blind users

<julierawe> julierawe Is browser support as robust in languages other than English?

<julierawe> Lisa I think we need to discuss language support as well as AT

<julierawe> Lisa For now the responsibility is on the author

<julierawe> Lisa This person is thinking you can automatic simplified text, but as far as I'm away, the AI isn't there

<julierawe> Lisa The assistive technology I was using was sometimes adding the wrong pictures

<julierawe> Lisa With cognitive disabilities in particular, when AI is wrong, it's much worse when the information is wrong

<julierawe> Lisa When is it good enough? Should we be trying to say if these assistive technologies get good enough, do you have to rely on which assistive technologies you're working on?

<julierawe> Rain I'm extremely hesitant to suggest anything that uses AI as a proposed answer

<julierawe> Rain It's still far too early to solve some of the challenges that we're facing

<Becca_Monteleone> +1 re: hesitancy with AI, particularly for simplification tasks.

<julierawe> Rain The AI itself, how do we know what it's been trained on? Does what it's been trained on meet the needs of the individual?

<Lisa> +1

<kirkwood> think some of concern is valid. one of the aI as potential (reviewed) ai. author approved. ai. ai warnings are important. surfacing where it is being used.

<JMcSorley> +1

<julierawe> Rain it can be leveraged with caution, it can be used in addition to what we've put in place

<julierawe> Rain When we think of AI in this context, there are newer studies that did some testing of some human-generated plain language vs AI-generated plain language

<kirkwood> most important thing is for reader to know WHEN it is used. and creating the structure to properly surface it.

<julierawe> Rain For text participants with cognitive disabilities, the human-generated plain language was easier to understand. But the original language was more understandabe than the AI-generated plain language

<julierawe> kirkwood Agree with Rain. One thing we can and should do is surface when it is used at all times. Having the proper language of how it's being used so the user can know AI is involved

<julierawe> I have to leave for 12pm meeting, thanks everyone

<Lisa> thanks julie

<kirkwood> is it “author reviewed” ai or not. to surface to be able to make their decision

<JMcSorley> Lisa: (1). AI could be a great tool for the author, but the author would need to review it to make sure that it works properly - kind of like using automatic captions and then editing them. (2). If people want to use automatically generated summaries, they need to notify people that it's been automatically generated because the summaries will only

<JMcSorley> work if tested.

<kirkwood> +1 to paragraph

<kirkwood> think Phil comment is valid, too

<kirkwood> “human-reviewed ai”

<kirkwood> +1 to issue paper

<Zakim> Rain, you wanted to add a thought on what that language should be