Re: Proposal: We need to identify whether a proposed SC applies broadly

Hi Jim,

Perhaps, but the language influences the navigation as well, and there are two other plain language SCs in the issues queue that appear to apply more broadly.

My point was that it would help to have a mechanism to target those at sites with a general-public audience.

That concept could also help with the ones that make use of 'critical features' and similar things which require UX knowledge & process that not all sites have.

Cheers,

-Alastair

________________________________________
From: Smith, Jim <smithjs@atos.net>
Sent: 24 March 2017 10:02
To: Alastair Campbell; kimberlee.dirks@thomsonreuters.com
Cc: w3c-wai-gl@w3.org; greggvan@umd.edu; lisa.seeman
Subject: RE: Proposal: We need to identify whether a proposed SC applies     broadly

All,

The scope of the plain language SC is as follows:

"Provide clear and simple language in instructions, labels, navigational elements, and error messages which require a response to continue,...."

The technical content of a journal is out of scope for this SC.


Regards,
Jim



-----Original Message-----
From: Alastair Campbell [mailto:acampbell@nomensa.com]
Sent: Friday, March 24, 2017 12:52 AM
To: kimberlee.dirks@thomsonreuters.com
Cc: w3c-wai-gl@w3.org; greggvan@umd.edu; lisa.seeman <lisa.seeman@zoho.com>
Subject: Re: Proposal: We need to identify whether a proposed SC applies broadly

Hi Kimberlee,

I think you have a point, but I'd like to re-frame it a bit. (And CCing Lisa as this is core to the usability/accessibility SC conundrum.)

It doesn't take long find examples of people with disabilities at the top of "professional" fields, such as Stephen Hawkins, Joseph Danowsky [1], Andrew Phillips [2], Richard Branson [3].

You said that:
> "I'm really struggling with how to apply some of the proposed SCs to *all* websites."

I agree with that, but it isn't professionalism as such, it is:
1) specialism, and/or
2) assumptions that authors know enough about the users.

As an example of specialism, one of our clients publishes Physics journals, the really esoteric stuff that I don't even slightly understand despite having done college level physics.

The publisher is just that, a publisher. They take in journal articles from Physics 'professionals' (I think mostly researchers & professors), and publish them. They provide guidance to authors, but we are struggling to get them to apply simple alt-text, there is no chance they will ever be able to change the content the authors provide. The source content is well structures (Latex), but even with a corpus of 'common words' from their own archive, I cannot see how the Plain language SC could be applied without making the content demonstrably worse for their specialist audience.
It is definitely unreasonable, and probably impossible.

The second point is when words like "critical" appear for features, tasks or services. There is an assumption that the designers/developers/testers actually *know* what is critical from the user's point of view. This runs into problems when:

- The organisation is fairly small, they may not have the resource for user-research.
- The organisation is large, and different teams (e.g. testers) may not have the access or time to understand the user-tasks enough to know what is critical.
- The interface has to offer many features (e.g. a word processor). There is the old truism that 80% of people only use 20% of the functionality - but each person uses a different 20%. You can't remove functionality without impacting your sales.
- Different users have different goals / tasks, and you don't have the ability/knowledge/resource to customise it for different audiences (regardless of accessibility).

Greg commented that:
> You can solve this problem by restricting the SC to only those types of content that it should apply to.

That is the WCAG 2.0 approach, but I don't think that works for criteria which require you to know what the critical features are. Also, for an SC like plain language how do you separate content by topic rather than interface?

I suggest we continue refining criteria in the WCAG 2.x approach, but rather than water-down criteria to the point they don't meet the user-need (or move them to AAA), I think it would be better to mark them as WCAG 2.1+ (or something, I'm not set on a term, just trying to convey the concept).

The criteria for "plus" would be specifically aimed at organisations which:
- Have a general public audience (e.g. Governments and companies catering to the general public), avoiding the specialist content issues.
- Have sufficient scale to do the sort of user-research / analytics required to know what the critical tasks/features are for their user-base.

It should allow criteria to be more "squishy", saying things like "where appropriate", "based on usability testing". etc. An SC could be A+, AA+, or perhaps we move it to the side of the A-AAA system so it is just +?

I think it also needs to be outside (or in some way separated) from 'conformance' as WCAG 2.0 defines it, because in order to test an SC in the plus category, you need to be part of the UX process. These are process check-points more than content aspects.

That's my preferred approach for getting out of the usability/accessibility conundrum pre-Silver.

Kind regards,

-Alastair

1] http://finance.yahoo.com/news/david-pogue-on-iphone-voiceover-163733668.html
2] https://en.wikipedia.org/wiki/Andrew_Phillips_(lawyer)
3] https://en.wikipedia.org/wiki/Richard_Branson

Received on Friday, 24 March 2017 13:01:54 UTC