Can Assistive Technology Make a Website Accessible?

What is Assistive Technology?

Assistive technology or adaptive technology (AT) is an umbrella term that includes assistive, adaptive, and rehabilitative devices for people with disabilities and also includes the process used in selecting, locating, and using them. AT promotes greater independence by enabling people to perform tasks that they were formerly unable to accomplish, or had great difficulty accomplishing, by providing enhancements to, or changing methods of interacting with, the technology needed to accomplish such tasks. Wikipedia

In the context of this discussion, we’re specifically referring to hardware or software devices that provide assistance to users with disabilities when using ICT products and services. Typically when people think of assistive technology they often think of screen readers or screen magnifiers, though there are a wide array of assistive technologies in use. On the web, assistive technologies are employed by users to help them in navigating, interacting with, and understanding content and features.

One of the common misconceptions people tend to have about assistive technologies is that assistive technologies can “make things accessible”. The assumption in such cases seems based on an incomplete understanding of how assistive technologies work – specifically the belief that features like all of the special keystrokes available for JAWS users to access certain types of content will somehow bridge the gaps in a system’s accessibility. The reality is that assistive technology types, features, and quality can vary quite significantly. Even in the best case scenario, assistive technologies can only render information to the user that is available from the system. In other words, if a page has a form on it, the ability of the assistive technology to provide information to the user about the controls in the form is dependent upon the form containing proper markup to convey that information. For instance, if form controls don’t have labels then what is an assistive technology going to do, invent one? In some cases that’s exactly what happens. JAWS will “guess” at form controls for text inputs, with limited success. This guessing about things like missing labels is rare in assistive technologies as a whole and is not consistent enough or reliable enough to count on it overcoming deficiencies in the system. The only way to ensure a system works properly with assistive technologies is to build the system right. In a web context, that means providing the most appropriate markup for the content and properly conveying structure, relationships and proper name, state, role, and values for controls. In other words, assistive technologies require proper web development in order for the user to truly be successful in using a system. This is exactly the opposite of what seems to drive the sentiment that assistive technologies can make things accessible. Assistive technologies can not make a system accessible.

The Rise of Pseudo-Assistive Technologies

Recently there have been discussions surrounding what I call “pseudo-assistive technologies”. For reference, I’m speaking specifically of products like ReadSpeaker, eSSENTIAL Accessibility, and BrowseAloud. In the strictest semantic interpretation, one could argue that they’re still assistive technologies with the primary difference being that they are offered on a per site basis and provide text-to-speech functionality to read the site’s content aloud. These products are actually not new and some of them are nearly a decade old. Despite this, the criticisms levied against such products are the same today as they were almost a decade ago. The problem I have, personally, with such products is twofold: first is what they’re capable of doing and the second is how they are marketed.

On a technical level these products essentially all work the same: They read the content of the page via text-to-speech. Though the features of each product differ slightly, the end result is the same: text becomes read aloud. The value you place upon the text-to-speech capability I guess depends upon your perspective. I will concede that there may be instances where this is useful. For example, I sometimes turn on ReadAloud on long PDF documents. To this end, these products are potentially useful for people with reading disorders or attention disorders. But here’s the thing: People who require text-to-speech in order to gain access to content will need it on all websites and, indeed, on all software applications they use, not just their browser.  This is important to understand: These per-site text-to-speech services are, at best, a mere convenience for persons who prefer to hear the content – not a legitimate assistance for those who need  to hear the content. This takes me to my second concern.

Deceptive Marketing

The technical limitations of these products are, in my opinion, not enough to warrant the rather severe negative reaction they often receive. To be honest, there are a lot of products out there – even assistive technologies – which don’t really do much or at least don’t do it well. If this were the only issue I had with these products, it would hardly warrant a blog post, especially when others have covered the issue well already. What I take issue with is the highly deceptive marketing of these products. Try these quotes on for size (emphases mine) which come directly from the materials on these vendors’ sites:

“Effectively, it ensure that anyone with a physical disability can easily access any website that carries this mark.” (eSSENTIAL Accessibility)

“eSSENTIAL Accessibility allows any user to overcome any physical limitation to access a web site.” (eSSENTIAL Accessibility user’s guide)

“Instituto de Mayores y Servicios Sociales (Imserso) and 22 of its websites integrate ReadSpeaker’s industry-leading online text-to-speech service to provide better web accessibility to its users.” (ReadSpeaker)

These claims are demonstrably false for the exact reasons I specified earlier in this post. On March 12, 2012, Denis Boudreau performed an audit (PDF) on a site which uses eSSENTIAL accessibility. Here are a few of the things he found:

  • Missing, incomplete, or uninformative alt attributes for images, including actionable items
  • Missing label elements for form fields
  • Poor color contrast
  • Poor keyboard accessibility and focus control

All of these items can present significant and, in some cases, insurmountable accessibility barriers for persons with disabilities regardless of what assistive technology is used.  To put it more succinctly: Not one of these issues can be addressed by software which merely reads text aloud.  Interactive features on these pages are inoperable by users with disabilities.  Even worse: The ReadSpeaker widget is, itself, inaccessible. In other words, the net outlook on accessibility for a site that carries this product is that the site becomes less accessible for featuring the item, not better.

Why this is bad for accessibility

Having a product that doesn’t do much is one thing. Claiming that it can do things that it can’t is something different altogether. These things are like the penis enlargement pills of accessibility. Some customers are so desperate and ignorant that they are almost eager to be duped. Just a little over a month ago, the Co-operators (the site Denis Boudreau assessed, by the way) issued a press release that proudly pronounced: The Co-operators website now more accessible merely by featuring a link to eSSENTIAL accessibility. Another press release issued just yesterday by the Ontario Medical Association says Almost one in five Canadians are living with disability. The OMA recognizes the importance of serving this demographic properly. But the really amazing one is this: Miratel Solutions Expands Their CSR Business Initiatives with Launch of Fully Accessible Website Clearly, these are organizations who have been led to believe that these services can remediate existing accessibility problems on their website. That is bad for accessibility. It is bad for users who need an accessible site because these products give organizations a false sense of security and the belief that they’ve done what it takes to serve people with disabilities.

There is only one place where accessibility happens: The code.  This goes for any assistive technology.  If the markup is used improperly, if the proper controls aren’t used in the proper ways, if the client-side scripting doesn’t manage focus properly and if the controls aren’t discoverable and actionable by keyboard, no assistive technology product, service, or device will matter.  The code is where accessibility happens and so long as companies’ budgets are being diverted to ineffective products, these budget dollars are not being used on things that matter, like training developers and remediating their sites.

If you are interested in learning about the next generation in Web Accessibility Testing, sign up for the release of Tenon.io
If you or your organization need help with accessibility consulting, strategy, or accessible web development, email me directly at karl@karlgroves.com or call me at +1 443-875-7343. Download Resume [MS Word]

5 Comments

  • niclasbergstrom
    Posted April 20, 2012 at 7:40 am   Permalink

    Hi Karl,

    Interesting post. And no, technology cannot make inaccessible websites accessible. Developers can, using the right techniques.

    I have a few comments on your text in regard of the ReadSpeaker services and market communication though.

    Speaking for ReadSpeaker, we always try to communicate that ReadSpeaker does not make an inaccessible website accessible. We make sure that all our communication is fully transparent on this essential point. We can’t always control how the customer website communicates on the addition of our feature though. We systematically educate our customers on what ReadSpeaker delivers as a service but can’t impose how they decide to market ReadSpeaker. When you refer to the Imserso quote, you don’t paint the whole picture since we do specify that “by speech-enabling its websites, Imserso reaches out to a wider population thereby ensuring greater access to its web content” We never say that adding ReadSpeaker to a website makes it become accessible. You seem to use a deceptive technique in offering incomplete information to your readers ;-).

    However, adding a speech functionality adds accessibility to the text content to a lot of people. The ReadSpeaker usage on the thousands of websites we speech-enable is in the order of several millions per month. We get a lot of positive feedback from both website owners and the users of their sites that adding audio to websites helps a lot of people. It is also a good way to introduce the text-to-speech technology to users that are not used to it. When it comes to accessibility in general, I would find it odd that anyone could claim that adding ReadSpeaker to a website decreases the accessibility to the website!?!? People that have severe reading (or visual) disabilities obviously need their own AT to be able to use the computer at all, and these are not the people we target. We’re not trying to push refrigerators to Eskimos here.

    Looking at the success Criterion on 3.1.5 of the WCAG 2 http://www.w3.org/TR/2012/NOTE-WCAG20-TECHS-20120103/G79 for example, the W3C mentions server based speech enabling of online content as something that increase the accessibility of the text itself. Unlike the eSSENTIAL Accessibility and the BrowseAloud solutions, ReadSpeaker is an integrated function in the actual website.

    I am not sure about what you are referring to when you mention “the ReadSpeaker widget” that you claim not to be accessible. As we have thousands of websites and we have a number of ReadSpeaker services (some pretty old versions still implemented), it would be great if you could share information

    • karlgroves
      Posted April 22, 2012 at 7:15 am   Permalink

      Niclas, thanks for taking the time to reply. Also, thank you for clearly stating your understanding that technology cannot make inaccessible websites accessible.

      You state “Looking at the success criterion on 3.1.5…” but what you linked to was actually a General Technique. Success Criterion are normative, whereas Techniques are informative, and the difference is important. As it states on the Technique page you linked to, “If this is a sufficient technique for a success criterion, failing this test procedure does not necessarily mean that the success criterion has not been satisfied in some other way, only that this technique has not been successfully implemented and can not be used to claim conformance.” Additionally, it bears mentioning that 3.1.5 is a Level AAA Success Criterion.

      Now, onto the accessibility of the ReadSpeaker Widget: I’ve checked two sites which are ReadSpeaker customers: http://www.mcu.es/ and http://www.jnj.com/connect/news. In the first case, the MCU website, the item opens in a new window and is not keyboard accessible. The reason it is not keyboard accessible is because it is within an OBJECT tag and the ‘wmode’ parameter is set to ‘transparent’. What this means is that no matter how accessible the item is underneath it, you won’t be able to gain focus on any of the controls. Consequently, keyboard focus cycles between the ReadSpeaker logo, the close window link and the MP3 link – never getting any focus on any item in the ReadSpeaker widget.

      The second one I looked at is on the Johnson & Johnson site. The version on the Johnson & Johnson site is much better in that it resides in the same page and provides visual feedback on what words are being read. However, it is still not fully keyboard accessible. Once the widget is opened, the ‘Play’ button has focus and can be accessed via the keyboard. Once the play button is activated, only the ‘Pause’ and ‘Stop’ button get focus. Keyboard focus is not possible to any other controls, including the control to close the widget, the settings link, or the link labeled “No Sound?”. This is why I stated that net accessibility is diminished. By featuring this item, new accessibility problems are created.

  • Posted April 27, 2012 at 11:31 am   Permalink

    Hi Karl,

    Sam here from Texthelp, the company that develops BrowseAloud.

    Thank you for writing the blog. It was interesting reading and it’s good to have an open debate on the topic, which is why I want to respond to a few points.

    We absolutely agree with you and Niclas Bergstrom that assistive technologies – in their current form – cannot make a website accessible. A website owner must make their website accessible and W3C have created a good set of international standards to assist with this.

    We also know that there is no ‘one size fits all’ solution to website accessibility; web users have diverse needs. This is partly demonstrated in the WebAIM article “Cognitive Disabilities Part 2” (http://webaim.org/articles/cognitive/conceptualize/), which discusses the difficulty faced in agreeing on guidelines that can be applied to web content for individuals with cognitive disabilities.

    BrowseAloud does not make an inaccessible website accessible. It works best with fully accessible websites and we actively encourage accessible design.

    However, there is a correlation between disability and unemployment/low earnings. While you are correct in saying that people who require text-to-speech will need it on all websites and all applications, what we aim to do with BrowseAloud is to provide good quality assistive technology for people who may not be able to afford it.

    BrowseAloud is free for the end user and has the quality and features of an assistive technology product that might normally cost the user a few hundred pounds.

    You may be interested to learn that Texthelp is working on a number of initiatives that we hope will go a long way towards improving industry standards. For example, AIA (http://www.atia.org/i4a/pages/index.cfm?pageid=3743), GPII (http://gpii.net/node/108) and Cloud4All (http://raisingthefloor.org/cloud4all). We are deeply invested in and passionate about providing fair and equal access for all.

    I hope this gives you and your readers a clearer idea about what we do and I’m happy to discuss further :)

    • karlgroves
      Posted April 27, 2012 at 11:57 am   Permalink

      Sam, thanks for writing. I think I should also mention that I’ve not yet seen any evidence of BrowseAloud being marketed in the same way as, eSSENTIAL is.

  • Posted April 27, 2012 at 12:36 pm   Permalink

    I’m glad to hear it, Karl. There have been a few occasions over the past few years where someone in an organisation has misunderstood the technology, or misrepresented it as an accessibilty ‘fix’, but this is not what we do and we try to get this remedied as soon as possible.

    I had a chat about the blog with our founder Martin McKay and we really do welcome your observations and critique. It helps us to identify what areas we need to improve and communicate more effectively.

2 Trackbacks

Post a Comment