WebAIM - Web Accessibility In Mind

E-mail List Archives

Re: SC 3.3.2: Labels or Instructions (Level A) violation?

for

From: Brooks Newton
Date: Apr 2, 2024 2:32PM

In my opinion, limiting the interpretation of SC 3.3.2 Labels or
Instructions to a narrowly scoped rule for text inputs discriminates
against a broad range of people with disabilities. I know I'm not the only
professional in this field who has run into scenarios where users of screen
reader software get better instructions for operating page content, forms,
and other types of interactive functionality than users with disabilities
who don't use AT. And, I don't think screen reader users will necessarily
get a clear representation of how a complex custom interface works by
simply providing programmatic indications at the component level without
adequately explaining how all of the integral parts of the interface work
together (also known as "instructions.")

The text of each Success Criterion (SC) in WCAG is normative. This
normative text should be able to stand on its own over time. Text in the
Understanding documents that accompany each SC is informative, and that
text will likely be edited over and over to suit the needs of changing
technologies, implementations, design sensibilities and even the politics
of the day.

For SC 3.3.2 Labels or Instructions, the normative SC text is as follows:
"Labels or instructions are provided when content requires user input."

I take a broader perspective on "user input" as do some in this forum
project. I think of "user input" as whenever users are required to
interact with the page/screen.

There are a lot of interactive page features designed by folks of all
abilities, where adequate onscreen instructions or labels are left out
because the designer simply doesn't recognize that some users with
disabilities (memory difficulties and other cognitive deficits, low-vision,
etc.) aren't going to be able to access content without some onscreen
help. Notice I didn't say these users with disabilities aren't going to
have an optimal user experience. What I said is that some users with
disabilities "aren't going to be able to access the content" due to the
omission of onscreen instructions and labels that are critical to
understanding page interactivity.

When a person compartmentalizes a Success Criterion by interpreting it with
the narrowest possible meaning, they may be favoring some disability groups
over others. I don't think that works well for the WCAG standard or the
broad community it hopes to serve.

Here's an example: If you have to have a full view of the screen to know
how all of the component parts of a custom interface work together, then
that application will likely need onscreen instructions to provide access,
not just enhanced usability, to people with certain disabilities.
Beneficiaries who may need onscreen instructions in this scenario include
low-vision users who may not get the benefit of understanding cause and
effect of the interface's component parts when zoomed in tight to a single
control. The interface's structure, logical dependencies, even component
roles may be ambiguous to users with disabilities if onscreen instructions
and labels available to all are not included in the design.

The page designer might have thought that instructions for operating the
interface were unnecessary and maybe even junked up the screen that added
to the page's "cognitive load." Since when did well-written, succinct
instructions make operating a web page more difficult? Plus, designers can
use progressive disclosure techniques to lighten the cognitive load for
users who don't need onscreen instructions.

Burying labels and instructions in the page code hoping that users with all
disabilities impacted will have the appropriate assistive technology (AT)
to uncover critical interface information is, in my opinion, a failure to
conform to S.C. 3.3.2 Labels or Instructions.

I'm not OK with just letting this narrow interpretation of WCAG SC 3.3.2
fly by on the discussion thread without commenting. Of course, this is my
opinion, and I may be wrong. But I don't think so, and I'm willing to hash
this out with anyone who wants to discuss.

Brooks Newton


On Mon, Mar 25, 2024 at 1:07 PM Patrick H. Lauke < <EMAIL REMOVED> >
wrote:

> 3.3.2 Labels or Instructions is aimed at controls that "require user
> input" - e.g. text inputs and similar. It's not a generic "any type of
> control that could do with instructions".
>
> So technically, I'd say this doesn't fail. Though of course it makes for
> bad usability if the keyboard controls aren't explained anywhere.
>
> P
> --
> Patrick H. Lauke
>
> * https://www.splintered.co.uk/
> * https://github.com/patrickhlauke
> * https://flickr.com/photos/redux/
> * https://mastodon.social/@patrick_h_lauke
>
> > > > >