3. Demands, Requirements and Use Cases

Accessibility within a learning domain is a many faceted problem that includes many different technologies and different systems. We also need to address a huge amount of different user needs. Reference to JTC1-SWGA- user needs summary document. And when we choose a holistic approach the complexity could be vast, but to try to illustrate the problem, we will choose one rather simple usecase from a actual learning situation and use this to point to where we will face some accessibility needs.

The use case we will use is the following one:

"There is a voice conference discussion going on, one person with hearing disabilities that prefer to participate in discussions using text enters the conference."

I don't see the point of introducing the use case here - you are not referring to it in the rest of this section! (Tore)

Unique requirements for Learning Education and Training

A basic accessibility requirements for Learning, Education and Training is that everyone should be an equal active member of the learning experience, with the possibility to participate and take part in the learning experience. Therefore is it important that the systems and resources have the capability to adapt to the users needs, and to change the mode of operation so that the learners have a barrier free access to the learning. Further the systems must have the possibility to provide alternative representation of content so that the learners have the opportunity to understand the content that is a part of the curriculum.

Said before? What are we getting at? What's unique (ref the heading) about this - in the context of accessibility?

In learning we will have all possible combinations of using, assimilating, sharing and creating information and knowledge, this multi discipline process will also be a process that involves many different learning technologies that have to support and scaffold the using, assimilation, sharing and creating process. To isolate out a single technology that does this is impossible - it has to be a combination of technologies, a combination of tools and methods - that all should ensure a barrier free access to the learning process not discriminating any users or usergroups.

Accessible tools

There exists a lot of accessibility support in operating systems and in the computer itself.

"in the computer itself" = programs that run on top of theoperating system? And below: "This tool?" This section has to be cleaned up! (Tore)

This tools are of a more general nature and not specifically addressing learning technologies. This tools are there for making the computer and the usage of the computer accessible, and therefore of great importance to learning technology as well.

Many of these tools are addressing the needs of the blind users, such as screen readers,  speech support, screen zoom, and the possibility to adjust for contrast.

When developing learning content and learning technologies the development should facilitate and support the needs of this technologies. Such as providing tagging in the web documents to be supported by the screen readers.

Predictability

To ensure accessibility one of the most important factors is the predictability. The more familiar your local environment and the cultural conventions are the more accessible the situation is for you. The same goes for learning technologies and learning content.

Accessibility to content could be separated into three different categories, syntactical, semantical and procedural accessibility.

The most common accessibility focus is on the syntactical accessibility, that includes how information is encoded and what kind of meta tagging that is supported. The WAI-WCAG recommendations are addressing the syntactical aspects. One of the major advantages of syntactical accessibility is with the ease this could be automatically tested by different tools and systems, and then more or less automatically "certify" that the information is accessible. This is done today by a number of organisations for rating different web-sites with regard to accessibility (http://www.eiao.net/) and one in Norwegian (http://www.norge.no/kvalitet/kvalitet2005/liste.asp), providing a set of "stars" for how "accessible" a site is.

One of the "problems" with such automatic testing of sites is that it does not necessary ensure that the site is accessible, because the site is created to satisfy the tests, and not to satisfy the needs of the users. One short example: many web sites uses images to control the layout and design of the page - one of the recommendations in the WAI-WCAG is that all images should have an alternative representation <IMG src="empty.gif" ALT="image used for design"/>. The question to ask is if the information presented in the ALT attribute is of any meaning to the blind user?

Semantical accessibility is about the possibility to make a useful meaning out of the information presented. The content should be adjusted to the readers skills and their level of knowledge. The more advanced the reader the more advanced the language in the content should be. It is important to observe that writing for the web is different than writing for a paper medium.

You cannot do an automatic test of semantic accessibility. The best way of testing a site for semantical accessibility is with real users from the targeted audience. This is a time-consuming and expensive, and would most likely not be done by most learning content providers. But it is important that the content developers have this in mind when they develop content. If the content is of a high level of difficulty, they should consider providing some alternatives of the content, i.e. provide some thesauri possibilities for difficult words.

The third category of accessibility is procedural accessibility. Procedural accessibility is about how you sequence and perform different tasks, and how things are organised. The more consequently this is done the more accessible is the information technology or content. When looking at different learning resources, we easily find that many of the learning resources are following a different kind of "logic" in the sequencing of events. It differs how you move from one place to another, where navigation structures are located, and where content and scaffolding information are placed on the page. As a consequence, the user of the learning resources need to learn a new design, a new navigational structure, and a new interaction scheme for each new learning resource. If all learning resources had the same structure and "design" the learner would use less time learning the design, and more time on learning.  To try an analogy, everyone can jump into any car and within a short time start driving. This is because the layout of the car, and the design of the car is consistent, i.e the accelerator is to the right. We need to reach the same level of interoperability with regard to learning technology as well.

Procedural accessibility is of great importance with regard to assessments and how the learner communicates with the systems, how you send information/documents and how that is acknowledged. Today the sequencing of events, navigation structure, and location of elements on a page is up to the developer of the system to decide. As a consequence there are great variety and a lot of new interaction schemes to learn.

Transformations vs. alternatives

Many content driven information systems today have a XML source. And using XML as a source have several advantages, one of them is that it is easily to transform XML into other markup languages such as HTML or DAISY. You could also transform the document into several different flavors of HTML to satisfy the different needs of the users. Even more media rich systems using FLASH as a front-end have XML documents as a source, and could therefore easily create accessible HTML representations of the learning content. This is one way of providing alternatives to text. But when we have moving pictures or audio only, these representations are impossible to transform. Alternative representations should therefore be provided, either with captions or as a transcript, or a combination of both.

Graceful Degradation

In conversations with users with different sets of disabilities they say that the most important is to get access to the information in any electronic format. They are quite experienced in interpreting the information - and as they say "we do not have any interest in lowering the learning experience for the average user". Based on this the systems need to degrade the learning experience gracefully - so that the normal average user still have a good learning experience, and so that the alternative or the transformed version for the learner with a specific need is not destroying the learning experience for the average user. Even if the system have to satisfy users with special needs, it should not be necessary to design all information and content based on that, but rather think of how could the information best fit for a as huge community as possible and how could the learning experience be of a as high quality as possible and at the same time satisfy other learners needs as well.