WebAIM - Web Accessibility In Mind

E-mail List Archives

Re: Where Does The Idea Screen Reader Users Use Tab for Main Navigation Come From?

for

From: Jonathan Avila
Date: Oct 11, 2018 6:02PM

Users I've worked with who learned JAWS before the virtual cursor or perhaps before the expansion of quick navigation keys were introduced tend to stick more to tabbing and the list of links and headings lists – this is understandable because this is all we had. When I started with JAWS 3.2 in 1998 you had to reformat the page into a single column and read down the page with the mouse pointer and you could access a list of links (If my memory is correct). You could also tab through form fields and links which was part of the standard browser behavior.



I'll extend this discussion to the use of the list of links features in screen readers. The WebAIM survey asked a question about how users find information on a page<https://webaim.org/projects/screenreadersurvey7/#finding>. Looking at the WebAIM data indicates that only 6.8% of users surveyed use the list of links feature to find things on a web page. Yet, if you talk to sighted usability folks who watch screen reader users they will say that the list of links features is almost always used and is critical. They will say that beginning users use this feature and that the WebAIM survey more broadly represents advanced expert users.



I'd like to make an assertion that behavior of some users may be related to training (perhaps by well-intended non-disabled folks who have a limited set of knowledge about assistive technology), experience, and limitations of older versions of technology. Some will say that we shouldn't change user behavior – and instead we should cater to user behavior instead. When I look at changes on the iPhone and in modern design and other areas I see people attempting to change user behavior for the better. When Microsoft Office came out with the ribbon every one was upset. And while some of you may still hate the ribbon – organizing contextual items out of the menus and into an area that is on-screen has really worked well for some users. A GUI while not accessible at first has been more accessible in time to users who couldn't grasp the command line interface. Apple took the touch interface of the iPhone to the next level and later made it accessible – all changing user behavior – but for the better. The touch interface which seemed inaccessible to many has been augmented with all sorts of accessibility features.



The list of links feature is useful – sometimes. I use it myself on occasion. Where it really excels is when you know the name of the link and you want to quickly jump to the link or activate the link. Where it is not useful is to understand the structure of the page and find something that you don't know what it's called. Take for instance a typical page with a store locator. If I've never been to the page before I could spend a lot of time reading through the links list looking for that link – or perhaps I can guess that maybe it starts with "s" for store and press "s" to move through all the links that start with "s". However, the name of the link could be "find store" or "locate store" in which case spending time digging through the list of links is actually slower than using other methods on the page such as checking the header or footer using landmarks or searching on the page.



Screen reader vendors also have a responsibility. Even after implementation of ARIA updates to the list of links feature lagged behind with no or limited support for use of ARIA to communicate the link name. That is now fixed – but even today the list of links feature in screen readers do not include the ability to speak the surrounding list item, table cell, sentence or paragraph. This is easily done in other situations like spell check where you can read the word in context with a command like JAWSKey+C.



In conclusion, I think it's important to ask why users do something a certain way. Is it because it's the way they learned many years ago or the way that someone without a disability taught them because that person didn't know the features and capabilities of modern assistive technology? Is the way still useful or effective? In what situations? Is our behavior dictated by limitations in assistive technology innovation? How can we drive better access to unfamiliar content and help users gain a overview of the page? The accessibility of controls and widgets of the page is just one aspect of accessibility. Access holistically to the page – understanding the core workflows and effectively and quickly being able to access the content at the same level of speed as others and becoming familiar with unfamiliar content is critical for inclusion. Some say that trying to change user behavior is like trying to boil the ocean. My feeling is that if there is a better way to do something an that way provides a better and more inclusive experience then we should be teaching others and spreading the word about these better methods of access. I'm not saying get rid of the old ways yet – but we can and should continually improve methods of access. I understand that logistically this is challenging – there is a lack of funding, training, updated assistive technology, inaccessible sites, etc. but those shouldn't prevent us from trying.



Ultimately landmarks, headings, and other structures along with ways to present these structure to the user and allow them to navigate them is key. We have made great progress in these areas – but I know there is much more that can be done – consider an automatic AI summary of the page, automatic movement to the part of the page you want to read, etc.. I'd ask that people consider the basic method of navigation tab, list of links, etc. and while useful and important to consider not assume that these are in fact the best way to navigate simply because most users know about them. Consider inclusive design that conveys a more wholistic view of the page rather than a link out of context and help educate all users.



Jonathan



Jonathan Avila, CPWA