Damian Sweeney - Instructional Designer, AIRport project, University of Melbourne
Much has been made of the legal implications of not creating accessible content for the web. While this is a legitimate motivation for educators to start creating accessible content, there is a more positive motivating factor available for those preparing online learning resources. Here we will show, through interviews with content providers as well as observations of the project's instructional designer, that preparing activities for an online environment with accessibility in mind has significant ancillary benefits. These include greater simplicity of design, higher quality text resources and, as shown here, a more creative approach to content generation. This can then frame accessibility requirements in a positive light for content providers.
"OK, today we'll be designing a web page for first year students. It has to be attractive, engaging, facilitate learning and respond creatively to information they provide."
When presented with such a task description, many content providers would be daunted by the task, but excited by the possibilities. They might be keen to try something a bit more creative for a change and already their minds are wandering in all sorts of intriguing directions. But there's always a catch:
"It also needs to be accessible. For our purposes that means all interactive activities have to operate on a text-only level. We can use images and rollovers, but these can't be central to the activity. No, Margo, we can't have pop up windows - it leads to confusion for lots of users. Sorry, Charles, that does mean we can't make use of Flash. Unfortunately, Jane, while we can use JavaScript, the idea you've mentioned won't be any good because it will refresh the screen without notifying the user. Any other questions?"
Perhaps this is the point where some people throw up their arms, others leave the room or bury their head in their hands. Perhaps, also, it doesn't need to be like this.
This paper explores attitudes to accessibility as perceived by content providers working on an interactive web site for undergraduate students. These content authors have been working for several months on interactive activities that must meet tight accessibility requirements. All support the notion that accessibility is important and that what we are trying to do is the right thing. Many have, however, expressed frustration that they are not able to do many things that are technically possible, but fall outside our accessibility guidelines.
We asked some of our content providers several questions about the content creation process and their attitude to the content they have developed. At this stage the respondents have all successfully produced at least two interactive activities that meet our accessibility requirements and are continuing to create them. Their frustration has diminished as they realise the possibilities that do exist and has produced surprising attitudes to the creative process.
AIRport (the Academic Interactive Resources portal) is a project of the Equity, Language and Learning Programs department of the University of Melbourne. It aims to provide a range of online academic literacy materials, including non-credit courses, to undergraduate students. This is a new mode of delivery for these materials and is designed to supplement existing programmes. The emphasis is on the interactive learning of academic skills online. These resources are expected to be of particular benefit to new students, international students and students from NESB. The AIRport will include three 'gates' leading students to:
We are aiming for an official launch at the beginning of the 2005 academic year.
As equity of access for students coming to the University is part of the remit of our department, we are seeking to be the standard bearers for accessible interactive content within the University. As such, we have set ourselves some high standards to aim for. This will become increasingly important for the University as it begins to meet its Access Melbourne targets where 20% of local (government supported) student places will be offered to students identified as having an education disadvantage.
The University has begun implementing a new set of templates that are rated as AAA compliant for accessibility (according to the WAI's Web Content Accessibility Guidelines 1.0). The AIRport project's design was completed before these templates were available, so we had neither the benefit nor the constraint of working with them.
For our purposes, accessibility means that anyone who can use a web browser should be able to use the web site without significant hindrance generated by the site itself. Note that this does not only include students with disabilities. Our rural and remote students are often connecting to the University at speeds of 14.4kbps and many choose not to download images. Providing a site that is text-oriented has become a priority for the project.
As we are providing content for all first-year students, the site also needs to be visually pleasant and usable for the vast majority of students not faced with obstacles to access. This increasingly includes students who are surfing University sites using their mobile phone or personal digital assistant (PDA).
To simplify the site design and to minimise accessibility issues, we have been proactive in our choice of technologies to deploy. Immediately excluded from the picture are mouse-dependent technologies like flash and browser-dependent technologies like client-side scripting (javascript). To maximise the reliance on text-based resources, server-side scripting (php) incorporating forms was used. This allowed a reasonable level of user interaction without immediate accessibility impediments.
As user authentication is a requirement for the site we have chosen to use cookies for the management of a variety of session-oriented data. Therefore, allowing cookies for the site is the only browser setting that needs to be enabled by the user. It would be possible to construct the site without using cookies, but as they are text-based files, are supported by all browsers and considerably reduce the complexity of writing code, it was an easy decision to keep them. Use is also made of the HTTP_REFERER header (which can be manipulated or disabled by some browsers), but this is not an essential part of the user experience.
We also deliberately separated style and content through the use of external stylesheets, thus addressing many of the problems of delivering content to different browsers. One issue we were confronted with was a page and site structure that was partly determined by the design of the site. This made compliance, particularly with old browsers, more difficult than it needed to be.
Another, perhaps more unusual, design decision was to force compliance with the XHTML Strict (1.0) standard. This proved to be enormously beneficial as many of the accessibility concerns were dealt with purely by conforming to Strict. Producing clean, clear and well-structured markup allowed for more time to be devoted to the content rather than worrying about accessibility at every turn.
It should be obvious, but it needs to be stated explicitly, that conforming to XHTML Strict guidelines is not the last word in accessibility compliance. There are many aspects of accessible design that can never be tested for with automated tools and reliance on such tools can give the wrong impression about the accessibility of a site. As with most issues related to web design, user testing is an important stage of the development process and should incorporate users who use a range of technical interfaces.
Software used for the creation of the pages was a combination of Macromedia Dreamweaver MX 2004 for Mac (mostly for templates and libraries) and the text editor vi. vi was preferred for page content because most of the tasks associated with creating the pages were in fact programming tasks and Dreamweaver is not as well suited to this.
Content providers in this instance consist mainly of learning skills advisors and transition officers. They are all new to producing these kinds of interactive materials for the web. Many have experience viewing similar web sites that we shall call "faux-interactive". These sites provide the opportunity for input from the user, but very little is done with this input. An example would be a page with a yes or no question, both of which link to pages with feedback. However, the feedback is identical except for a 'correct' or 'incorrect' message. Many also include rollovers or flash applications that provide a visually interesting user experience, but ultimately don't contribute anything to the learning experience of the user.
The majority of content providers have extensive experience with providing content for their field in a variety of forms including lectures, small group discussions, PowerPoint presentations, booklets, pamphlets, workshops and so on. Some have also provided content for web sites, but this content has always been static (non-interactive).
As part of the process of developing materials for AIRport, content providers were exposed to some of the accessibility requirements for the site through a demonstration of existing pages that conform to our requirements. Emphasis in these demonstrations was placed on the need for text-based representations of ideas and on the centrality of the learning experience to the activity. It is felt that the time and energy expended for these demonstration/training sessions is well worth the investment given the time saved in staff members creating activities that may never be realised.
There are still examples where accessibility requirements arise as an obstacle to design ideas. This usually occurs when the content providers are working on their first or second interactive activity. A question is often posed along the lines of "would it be possible to ...". Even this approach by content providers is a positive step as they are no longer arriving with a fully formed idea that is wrong. They are bringing an idea for an activity and are now seeking the best method to facilitate learning.
To gain some understanding of the various reactions to the accessibility requirements four content providers were given a questionnaire with the following questions:
For the first question, all respondents mentioned physical barriers, three mentioned visual difficulties and one mentioned cognitive impairments (acquired brain injury). One person mentioned modem speed and another referred to password protection (ie, restricting access). Two responses used the word usability when talking about disabled users interacting with the site.
For the second question, asking specifically about AIRport, only three people responded. They all mentioned 'mouselessness' as a requirement, two mentioned download times and one talked about creating text-based content. Text alternatives for images and avoiding Flash also appeared in one response.
So apart from the reference to password protection there seems to be a broad understanding of accessibility represented by this subset of content providers. When it comes to practically thinking about creating accessible content the emphasis on text-oriented materials comes through as a mind set being used by these authors.
In terms of benefits, as perceived by the content authors, two responses mentioned keeping the site simple and 'transparent'. One mentioned that it 'makes us put more thought into content' and another that 'although the process of production is more difficult, accessibility has encouraged creativeness'. Other answers referred to the benefits for learning outcomes from the materials because they considered 'multiple layers of learning' and reduced user frustration, thereby 'ensuring retention and engagement with the materials'.
The constraints, as perceived by content authors, revolved entirely around visual content. This is perhaps best summed up by one respondent who wrote:
"The main constraints are in use of visual instructional tools. For much of the content I am working on, the visual register has been an important teaching tool. It is possible to work around this, but the constraint causes a rethinking of a lot of the material."
In general, the responses demonstrate that the accessibility requirements are not perceived as a barrier to creating interactive activities by those creating the content. The benefits of considering accessibility at each stage include a more creative process for authors and simpler and clearer design process. Note that these benefits may not be transferable to web sites that are not interactive.
There have been times when we have compromised our accessibility requirements because of the nature of the content. One section of the site has been identified as providing poor accessibility, purely because the concepts being conveyed are visual in nature (e.g., mind-mapping). To exclude this type of valuable material would have reduced the user experience for most users. The pages are still usable and provide alternatives, but these do not deliver the same level of interactivity. Students who are unable to benefit from this section are given priority entry into the other, more accessible section. Content that is not amenable to accessible versions is a rarity though.
User testing was conducted with our major target audience, which is local and international undergraduate students (at this stage only Gate 1 content is ready for testing). Two testing phases used separate groups of students and different methodologies. The first group was observed as they used the web site, while the second cohort was asked to use the site in their own time and then respond with the online feedback form.
Both groups responded
positively to the content of the web site
describing it variously as
practical
, helpful for study
,
very helpful for
international students who are not familiar with
uni surroundings
and fun
.
The few negative comments tended to focus on the design aspects of the site which we are in the process of addressing prior to the launch.
In first semester 2005, we will be developing a section of the site for the University's Disability Liaison Unit and will be conducting extensive user testing for their client base. For this reason we have not yet undertaken user testing for people using adaptive technology.
From the perspective of an instructional designer, the requirement to have text-based interactive activities meant that much greater attention was paid to the quality of the content and the words being used. Many content providers would initially talk about rollovers and messages popping up as their contribution to the interactivity of the activity. When confronted with creating a Braille-friendly version of the same activity, attention would shift from presentational aspects to the learning experience afforded by the content. While this process led to the creation of quality content, it was not always popular.
However, as the activities started to materialise, many content providers have found that the requirement for text-based materials has focused their thinking. They find that they approach a task with a clarity they would not otherwise have had. There are times when this clarity has disappeared. These times, however, are usually when someone has seen an existing activity and is trying to adapt it to our needs. When starting from scratch, the development of the activities now proceeds quickly, efficiently and the actual text used is of a high standard.
The process of creating these interactive exercises accessibly is not always popular as content providers grapple with the constraints of a text-only site. This emphasises the fact that providing content for an interactive web site is necessarily a creative task and not a simple process of translation from text to web. High quality content is the result of accessibility requirements because creativity is required in the authoring process rather than in the presentation of the material.
So, if you want quality material on your interactive site, then mandate conformance to accessibility requirements. If you want content providers to be enthusiastic about working within accessibility constraints, show them first what is possible within those constraints.