WebAIM - Web Accessibility In Mind

E-mail List Archives

Re: Clarification on forms mode of screen readers


From: Birkir R. Gunnarsson
Date: Jul 26, 2016 8:01AM


If you ar formally testing a webpage for accessibility, I recommend
you run an automated accessibility testing tool, such as aXe, on it
The tool will catch and report issues such as when a label is not
provided for a form field. That is easier and more reliable than
relying on a screen reader to do that testing.

Then you can analyze the code behind the webpage in a couple of ways.
I wrote an article on how to inspect webpage code with a screen reader
and Firebug:
I used to test with the Wave toolbar from WebAIM as well, but I have
not found a good and easy way to use it with Chrome, and it is no
longer available in Firefox (I heard a rumour that it may be coming
back, I will be excited when it does).

On 7/26/16, Moore,Michael (Accessibility) (HHSC)
< <EMAIL REMOVED> > wrote:
> If you are visually impaired and using a screen reader to test whether the
> form labels are properly associated with the form controls I recommend using
> NVDA to perform those tests. NVDA does not guess like JAWS does.
> Mike Moore
> Accessibility Coordinator
> Texas Health and Human Services Commission
> Civil Rights Office
> (512) 438-3431 (Office)
> -----Original Message-----
> From: WebAIM-Forum [mailto: <EMAIL REMOVED> ] On Behalf
> Of Vemaarapu Venkatesh
> Sent: Tuesday, July 26, 2016 7:12 AM
> Subject: Re: [WebAIM] Clarification on forms mode of screen readers
> Hello all, Greetings
> On the way to understand the interaction modes in JAWS, I got encountered
> with few more uncertainities
> From previous messages I got to know that screen readers will switch the
> modes automatically and if that is the case no need to switch manually which
> is happening perfectly.
> Now as a screen reader user, when I am testing application using JAWS, it
> starts guessing the things like the form lables and announce them even
> though they are not associated. So, how the screen reader user will get to
> know that the lable is programmatically labelled or not. Yes, if it is
> labelled the label itself becomes clickable but how a screen reader user
> will know as he can't click and JAWS even not announcing clickable when
> focused to label text in virtual mode.
> What I understood is if we turn off "virtual pc cursor" manually, JAWS is
> rendering everything perfectly and I didn't find guessing behaviour in off
> mode.
> The same with "select country of residence" combobox example provided by me
> in earlier messages.
> My question is, whether I have to test all form elements by manually turning
> off the virtual mode because I can't able to differentiate between
> programmatically labelled elements and unlabelled elements in virtual mode.
> But in NVDA I didn't find any such differences, form elements are similar in
> both the modes. But I am finding difference in behaviour of JAWS not only in
> form elements but also in few other contexts when I am manually switching
> between virtual pc cursor on/ off.
> Hope my concern is clear and clarify me about how to proceed my testing with
> Regards,
> venkatesh
> > > http://webaim.org/discussion/archives
> > > > > >

Work hard. Have fun. Make history.