Research Proposal: Accessibility Support for Non-HTML Web Technologies

Written by Mark Magennis Monday, 22 February 2010

This proposal is for research into the state of 'accessibility support' for key web technologies that present significant barriers for web users with disabilities. Examples include PDF and Flash. We're looking to do something significant in this area, either alone in Ireland or in collaboration with others internationally. We are particularly interested in assisting a student who wishes to do an MSc by research and is looking for a challenging, interesting and socially important research project. We are also keen to share with others within the international accessibility community who are working on this issue and who would like to combine knowledge, trade opinions and perhaps work together to further refine our proposed approach.

What is 'accessibility support'?

Accessibility support is a concept that has been introduced by the Web Accessibility Initiative (W3C WAI) in version 2 of the Web Content Accessibility Guidelines (WCAG 2.0). Its purpose is to extend the guidelines to cover non-HTML technologies such as JavaScript, Flash and PDF. It provides a mechanism for web developers and website owners to determine whether a given technology can be used for a particular purpose, i.e. to present a particular type of web content or functionality, in a way that will be accessible to people with disabilities.

The basic concept is that a technology (such as HTML, Flash or PDF) should only be used for a given purpose (such as presenting an interactive form) if there is 'sufficient' support for this by the access technologies used by the user population. For example, HTML is accessibility supported for interactive forms because the access technologies used by people with disabilities can successfully read and interact with HTML forms (assuming the forms are correctly marked up in line with WCAG). But interactive forms in PDF (even when properly tagged) can only be read and filled in by the latest access technologies. Few users have or can afford these latest technologies, so the use of PDF for interactive forms cannot currently be considered accessibility supported for a general web audience.

So accessibility support operates at a high-level to rule out inappropriate uses of technologies. It should be possible to create a matrix of technologies against supported uses as a guide for web developers and website owners. That would be the first step of this proposed research.

In practice, however, accessibility support is quite a complex issue. There is no general agreement about what constitutes 'sufficient' support and levels of support are highly context dependent. It depends not only what the content technologies and access technologies are capable of, but also how reliable they are, how much the access technologies cost and to what extent users with disabilities can be reasonably expected to acquire them and develop sufficient skills to use them. This will differ from one user population to another. In the WCAG 2.0 documentation, WAI leaves these judgements up to "the community and to entities closer to each situation that set requirements for an organization, purchase, community, etc.". So far the community has done little to address these issues in a structured way that would lead to authoritative and coherent guidance for the entities that set requirements. The aim of this proposed research is to do just that.

Why this research is important

Access to online tools and content is now essential for social inclusion. The barriers presented by inaccessible websites can cause severe disadvantages for people with disabilities in their education, in employment and socially. Accessibility support for PDF and Flash in particular is a hugely important current issue for users with disabilities because:

PDFs are ubiquitous. On a typical Irish local government website, over 99% of all the information is contained within PDFs, not in HTML.
The use of Flash on websites is increasing rapidly. It is often used for essential functional parts of the user interface such as video players and site navigation. Some important sites or online tools such as the BBC iPlayer are made entirely in Flash.
Many people with disabilities have quite old access technologies. For example, NCBI estimates that over 50% of Irish users of the JAWS screen reader have versions 5 to 7. The latest JAWS is version 11. Upgrades or replacements are often prohibitively expensive, especially since most people with disabilities are unemployed.
Website owners who are committed to accessibility are often unsure what to do about PDFs and Flash. They know accessibility issues have always existed with these formats but they also know a lot of work has been done by Adobe and access technology vendors to overcome this. Now they're wondering "can we use it now that it has 'accessibility features' built in? If not, what can we use instead? And when will access technologies have caught up so that we can use PDF or Flash in the knowledge that they are accessibility supported?".
Even when accessibility features exist within these technologies, website developers and owners are often not aware of them or don't have sufficient skills to exploit them correctly.

Although we are currently concerned most about PDF, Flash and to a lesser extent Microsoft Word documents, the research can be extended to any other technologies that are being introduced or are on the horizon and which may have as big an impact and create as many problems for people with disabilities.

Outline of the proposed research

The first stage of the research would be to systematically and comprehensively measure the abilities of different access technologies for interacting with different types of content and functionality in different formats. Access technologies would include multiple versions of screen readers - JAWS, WindowEyes, NVDA and Voiceover for example - and screen magnifiers with different operating systems and browsers. Content and functionality would include structured text, interactive forms, data tables, audiovisual media and rich interactive applications. Formats would include PDF, Flash and Microsoft Word. This would allow the creation of a matrix of technologies against uses against levels of support.

The next stage would be to carry out a survey of web users with disabilities to ascertain what access technologies they have available to them and what levels of skills they have in using them. This should be done for different user communities. As a start, we would like this to be done for the Irish vision impaired user community.

The combined results of these two studies could then be used to estimate, for specific uses of specific technologies, what proportion of users with disabilities in the community would be excluded by reliance on that use. It would also provide evidence on which to reason about the extent to which users within specific user populations can be reasonably expected to have sufficient access technologies and sufficient skills.

In order to propose a way forward, it might also be useful to survey the extent to which accessibility features in online technologies are known about and correctly exploited by website developers and owners.

Call for participation

This constitutes a challenging but interesting and socially important research project with international impact. It would be an ideal subject for an MSc by research or as part of a broader PhD research. It would make a significant contribution to the knowledge of real world online accessibility and would produce highly publishable results.

The NCBI Centre for Inclusive Technology (CFIT) would be able to facilitate this research by providing the following supports and inputs:

Training in access technologies and web accessibility.
Access to users with disabilities for studies and evaluations.
User testing facilities if required.
External supervision and subject expertise.
Assistance with writing and publishing research papers.
Any other practical assistance necessary.

The researcher would benefit from close contact with a centre of excellence in universal design and digital technologies.

We issue this proposal as a call for assistance in finding the right individual to take on this project and identifying a source of funding. We would welcome any comments, criticisms and collaborations in order to refine this proposal and get this work done.

Add your comment

Your name:
Your email:
Your website:
Subject:
Comment (you may use HTML tags here):

Comments (9)

hi,
if anyone wants to take up MSc/PhD or contribute in any capacity, how should one go about it?
would love to know that.
thanks

I don't know where MSc research students get their funding from Shriang. That's one of the things we're looking for input on. Hopefully, someone in the academic community will be attracted to this idea and will steer a student towards it and help organize the funding, or at least give pointers to where to look.

Most of the research proposal talks about Windows based programs and screen readers. Linux is increasing in number of users including users with disabilities. So, it would be great to have Linux included with Open Office support, as well as flash and PDF support. The screen reader Orca provides access to the Linux OS and the web:
http://live.gnome.org/Orca
Orca also provides magnification, not just speech and Braille.

If we're talking about a single research MSc, my guess is that a narrow focus on just PDF, Flash and Windows based screen readers might be as much as can be achieved, especially since we would want to cover a very wide range of content types, from structured text, through forms and data to audiovisual content. If we are able to collaborate with others, then we might collectively be able to cover a lot more technologies. Our initial primary focus will be whatever we perceive to be causing the most problems for our user community here in Ireland. Right now, we don't perceive a great amount of Linux use, at least among the Irish vision impaired community, but we would look carefully into that to determine the right focus.

We would, however, hope to define a methodology that could be repeated for any technology - different web content technologies as well as different user agents and platforms. Then if we extend this research outside of Ireland or collaborate with others internationally, a new scoping exercise might result in a different focus, including Linux. But in the end, it's all about hitting the major problems first which is why we're looking at Flash for example rather than Silverlight and why we'll probably (not certainly), start by looking at Windows, PDF, perhaps MS Word, rather than Linux, Open Office, etc. I take your point that it would be great to cover all technologies if possible.

I would love to help with the PDF aspect of this. I've conducted research regarding PDF accessibility for the past couple of years and would love to help and learn more about this.

Thanks!

There are two purposes for PDF and unfortunately they conflict. A viewer application so people can view/print typical documents without owning that application (Accessibility in mind). The other is inexpensive reproduction of print materials (various paper sizes) often beyond the scale of accessibility such as CAD, blue prints, multi-column layouts, complicated tables/graphics (often beyond today's accessibility capability). Side note: The simple scanning of pages as images (as many major copiers do, text included) complicates & hinders such efforts as well. Separating these two product purposes hasn't not been a priority. History shows you can not satisfy both camps with a single product.

Hi,

I develop resources for reading with low vision that do not just rely on magnification. These include tag based navigation, apropriate and granular typographic enhancement and usability.

My work is for reading. Numerous digital libraries depend on PDF for document storage and delivery. So, the move to PDF and Flash directly threatens literacy of people with low vision.

For many zoom / magnification is not enough. Granular control is needed. No PDF reader or zoom product provides releif for this group.

I am eager to help.

Wayne Dick, PhD
Professor, Computer Engineering and Computer Science, Long Beach CA.

Firstly, I have to commend the original author of this discussion/suggestion. I've read countless postings online and rarely have I come across something that is so well thought out.

Despite Adobes best efforts, their PDF/Flash technologies by design, will always be a major obsticle with respect to screen reading technology.
The issue lies not with the tools, but rather the additional layer of sophistication required to make the technology function with adaptive software.

In many aspects, asking publishers and designers to incorporate accessibility into PDF/Flash web media is almost the same as admitting that the technology isn't accessible and there by needs to be retrofitted. As with most acomodations for anyone with a disability, it's been proven over and over again, that the most effective approach to make any product or technology truly accessible is to build reasonable access from the beginning. In this case, that means that the paradime needs to shift from designer to the publisher of the tools.

Specifically, only create objects and authoring tools which automatically structure the end product in an accessible format. Expecting authors or web managers to stay current on changing access technology, its limitations or strengths isn't a long term viable fix.

Having said this, I have to take exception with a point made in the original thread. Its been stated that this research isn't a critique of Adobe; why not. I suspect that if truly put to the test, most of the Adobe tools would fail even a simple accessability screening. For example, Dreamweaver CS4, touted by Adobe to be an accessible web page authoring tool is completely incorrect. Yes, much functionality probably exists within the application to be able to author accessible sites, but the authoring tool itself fails miserably when put to the test. Yet, I would wager that at least a small part of Adobes success in marketing this technology has at least a small part to play with the perception that Dreamweaver is accessible.

The answer? Clearly differentiate between products which are accessible and those which can be used to help improve accessibility of the finished produt. This, isn't splitting hairs as I'm sure some might hold.

@Donna, @Wayne: Thanks for your offer of help. I would love to involve you. As I see it, there will be two ways of involving people. First, the research project will have to start by gathering all the information and data that is currently available on the subject. Secondly, the researcher will have to collaborate with people like yourself who are already working on this issue. This could involve agreeing the distribution of testing tasks among various people, in a structured way, in order to reliably cover all the ground without duplicating effort. This would allow a lot more hard data to be produced than could be done by one person alone and would also allow for some controlled redundancy to help ensure accuracy. We'll be in touch!

yvComment v.1.22.0

Home