Task completion: Successfully completing tasks does not rely on users memorizing information presented in the current, or previous, user-interaction dialog steps.

I think if the word 'dialog" was removed here, it would reduce the prescriptive 'voice dialog' bias in this SC and allow it to have a broader application.

Description

In multi-step user-interaction dialogs, such as voice-menu systems...

I’m not sure why the focus has been reduced to voice-menu systems. That is obviously an example, but it’s not a typical web example. This is the first time I can recall non-web examples used for WCAG. I also don’t think it’s just voice systems that are presented serially; for example a questionnaire may ask a question per page, with a prior response determining current content.

See Voice Menu Systems issue paper...

Okay, I’m going to stop reviewing this SC at this point because I’m very confused about how a SC that is entirely non-web-related can be considered a WCAG SC. If WCAG has authorized COGA extending beyond the web, that’s interesting, and has ramifications on altering some of the highly web-centric language in parts of WCAG. But I would be surprised if this is actually the case.
On the topic of Voice menus, I also think this SC falls short, in that it does not introduce level A basic requirements for accessibility/usability (outlined in a number of documents such as Top Ten Voice Response Usability Violations). I would suggest this SC on memory resolution would then possibly build on that as a level AA SC.

I am concerned that this is potentially very broad. I'd like to see some cases where this would fail currently, and then to understand what it would take for them to pass this SC.

an example of a failure is a voice ML system (like a phone menu system) that sease
"press 3 for internal services" making the user remember a digit 3 whilst figuring out if
what thye want is an internal service
having it the other way around("for internal servces (pause) press 3" takes aways the memory requirement. the user can figuer it out and then hears the digit they need

Alternitivly reserving the 0 digit to get to a person, Or having the first option "to wieght for an operator press 0" also removes the memory rquirment

@lseeman I am interested in web pages-based failure examples. Can you provide?
(now seeing that is JF's question also)

(Replicating my survey comment)

The SC text refers to "completing tasks" but the testing talks about "user-interaction dialog step".

These are not the same thing, and a task is a very wide concept.

I think it should be narrowed to the actual focus of the description, how about:
"Using interactive controls does not rely on...".

I think that actually has the same breadth, but puts the focus on the web-content rather than the more vague concept of "task".

Can we merge part of accessible authetification. IE:

Users are not required to memorize information, character strings, and correct spellings; or perform calculations, such as including correctly identifying and entering numbers and letters from a character string; or speak, unless it is essential for the main function of a web site.

New wording could be

Task completion: Successfully completing tasks does not rely on users memorizing information character strings, and correct spellings; speak; or perform calculations, such as correctly identifying and entering numbers and letters from a character string or memorizing information presented in the current, or previous, user-interaction dialog steps; unless it is essential for the main function of a web site.

Is this new wording ok? Then we can get rid of capture from accessible authetification

I had a conversation with Lisa about this and it seems that at the core is the idea that there is a need to provide instructions that are delivered linearly (largely meaning by audio) in a manner that the user receives the differentiating information first and then the action after that.

For example, "press 3 for sales, press 4 for support" is a problem where "for sales, press 3; for support, press 4" is better for people with short term memory issues.

I think that if we are going to do something about this we need to provide that guidance. There would still be a the existing requirements for textual versions of audio, but if there was a web-based service that doesn't have a screen or is delivered via a phone, then this might be the only way to support the user need, and it may be useful if the user prefers audio over the text.

Here's a suggestion:
When presenting users information about multiple selectable options via audio, provide the details about each option before indicating the action that the user needs to take to select the option, or provide a way to get to assistance.

Then we would provide information in the understanding document that would provide examples.

What do people think? By making this about audio instructions, which certainly can be provided via the web. Thoughts?

I would also suggest renaming this "Audio Instructions" or something else similar.

In SC speak it would be something like:

When information about multiple selectable options is provided by audio recording, details about each option are spoken before the action that the user needs to take to select the option unless there are instructions or a way to get to assistance.

The big question is internationalization. In some languages that might be the opposite way for the best understanding, I don't know.

That's a good point, I'm not sure. @lseeman?

I suspect that there is actually something to the idea that people need to know whether the option is the right one before being told what to do about it.

There seem to be a couple of considerations here. I'm not worrying too much about wording here, just getting the ideas down:

1. Each step in a sequential process must contain the information necessary to allow a user to proceed (and so must not rely on memory from prior steps, and would benefit from a summary of info, or a mechanism for traversing the process.
2. In information provided by audio output, labels should proceed the activation mechanism

Are there others?

@mbgower
the best practice I think is to allow people to reach a human by pressing 0. however we need to allow other alternatives as well
@awkawk I am not sure why internationalization should be a problem. people have trouble remember in a digit while processing speech in any language. do you know a language where requiring rembering the digit is necessary? (Hebrew is a very different construct to english and i do know it works there)
@johnfoliot this might interest you (from the issue paper) "For example, the U.S. Telecommunications Act Section 255 Accessibility Guidelines [Section255] paragraph 1193.41 Input, control, and mechanical functions, clauses (g), (h) and (i) apply to cognitive disabilities and require that equipment should be operable without time-dependent controls, the ability to speak, and should be operable by persons with limited cognitive skills.
Technology" - in other words this is already a requirement by law, just people do not know it.

allow people to reach a human by pressing 0

@lseeman, As per my original comments back in February, why are these examples phone based instead of web-based?

U.S. Telecommunications Act Section 255 Accessibility Guidelines [Section255] paragraph 1193.41

255 got updated at the start of the year, so the numbering and text has all changed.

We can have just add a scope narrowing term

Task completion: In voice systems, successfully completing tasks does not rely on users memorizing information presented in the current, or previous, user-interaction dialog steps.
Note that the following requirements must apply: understandable language; clear structure and relationships; and clear and unambiguous separation of menu items, which must be level A in this scope. (This may be addressed by reiterating these requirements here to increase conformance.)
Exception: This success criterion does not apply if memorizing, or otherwise mentally processing information, is the primary purpose of the dialog step (e.g., a step in a game which deliberately tests the player's memory skills).

@lseeman

the best practice I think is to allow people to reach a human by pressing 0. however we need to allow other alternatives as well

So is this a third consideration?

or a mechanism is available to request direct assistance

As it stands the front text of the SC expresses a very general user need that applies in systems. The text "Successfully completing tasks does not rely on users memorizing information presented in the current, or previous, user-interaction dialog steps." has many different ways to satisfy it - and its very valid but if the way to satisfy it is as given in the techniques then in my opinion the SC needs scoping down to that context. Such as "When using voice audio systems successfully completing a task ...". I am new to writing SC text so I think maybe this narrow context is described in some structure into which the SC drops (I'm guessing), but as it stands we have a very general and broad "need" satisfied by a very narrow set of techniques. -andy