VLE Report | 7. Implementing accessibility in VLEs: survey results

7. Implementing accessibility in VLEs: survey results

This section maps the results of the survey and interview data gathered during this study, and compares them with existing research. The study utilises data from a total sample size of 57 (46 survey respondents and 11 interviewees). The data represented graphically in this section is drawn solely from responses to the online survey (which is set out in full in Appendix 1). Data from interviews is included in the textual analysis.

Represented in the online survey were:

23 higher education institutions (HEIs)
19 further education institutions (FEIs)
4 independent consultants.

Individual respondents were:

a majority of ILT specialists, with a range of job titles such as ‘e-learning adviser’, ‘instructional designer’ or ‘learning support adviser’ (n=37)
ICT specialists such as ‘C&IT support officer’ or ‘IT manager’ (n=5)
lecturers (n=4).

For ease of display and interpretation, some data is presented in the following sections as percentages, but it should be borne in mind that, with a sample size of 46 this may produce distortion. As discussed in section 2 on methodology, the small sample size of this study means results and conclusions are necessarily tentative. As will be evident, however, the results are consistent with other studies.

7.1 Overview of respondents' VLEs

Figure 2 is a bar chart showing the types of VLE in the study sample of 43, with Blackboard the most common, followed by webCT, Granada Learnwise, bespoke systems, Virtual Campus and miscellaneous others

Figure 2. VLEs by type/vendor

Figure 2 shows the relative market penetration of various virtual learning environments based on the data from the 43 survey respondents whose institutions had a VLE. The market leader (32%) across FE and HE is Blackboard, a US product. Second (24%) is WebCT, also a US product.

The VLE market is quite young, and so there are few sources against which to compare this data. However, Jenkins and colleagues (2001) showed Blackboard entering the VLE market around 2001, and quickly reaching a dominant position. The most extensive recent research, a national survey of MLEs in FE and HE (JISC/UCISA 2003), shows Blackboard as currently having the leading share of the VLE market at 33%, and WebCT second at 20%.

Next in this study is Granada Learnwise, a product from a UK company, with 13% of the market, followed by bespoke in-house VLEs at 12% and Virtual Campus from UK company TekniCAL at 11%. These figures are also similar to the national figures from JISC/UCISA (2003).

The survey showed differences between FE and HE, with FEIs showing a preference for Blackboard, and HEIs for WebCT. Granada Learnwise was used only by FEIs in our sample. Once again this finding is broadly consonant with findings from JISC/UCISA (2003).

The category of bespoke VLEs in Figure 2 includes Colloquia, Bodington Common and WOLF, examples of VLEs developed by individual academic institutions for their own use, some of which have subsequently been taken up by other institutions. The VLEs subsumed in the category ‘other’ are commercial products Doddle, First Class and Fretwell Downing. (See Appendix 2 for more information on individual VLEs.)

These comparisons imply that our sample, though small, is representative of the products currently used across UK FE and HE.

Figure 3 is a bar chart showing VLE uptake starting in higher education approximately 5 years ago, then starting in further education, peaking across both sectors between 1 and 3 years ago, and now slowing considerably as the market moves towards saturation

Figure 3. Uptake of VLEs over time (as at September 2003)

Figure 3 indicates uptake of VLEs over time in this sample, with a surge of uptake between 2000 and 2002, slowing in the last twelve months. With over 80% market penetration in FE and HE at present, this pattern would be broadly in line with Rogers’ S-curve pattern of adoption of new technologies (Rogers 1995).

7.2 Importance of accessibility

Respondents were asked:

As a criterion for your choice of VLE, was accessibility for disabled users:

not important
considered but not primary
of primary importance
a necessary pre-condition since the introduction of SENDA.
(Question 12)

Figure 4 is a bar chart indicating that 16% of 43 respondents did not consider accessibility important in their choice of VLE; 48% thought considered accessibility but did not think it was of primary importance; 36% thought it of primary importance or a necessary precondition for their purchase because of SENDA legislation

Figure 4. Relative importance of accessibility in choice of VLE

Reponses are shown in Figure 4. Fifteen per cent of respondents cited the accessibility of a VLE as a necessary precondition for their purchase/development, but the same percentage cited VLE accessibility as ‘not important’. All but one of those considering accessibility to be unimportant had chosen VLEs after the introduction of the SENDA. The largest group (48%) of respondents cited accessibility as a consideration, but not a primary one.

Clearly this data is open to a number of interpretations. It could be that because all the main commercial developers claim to comply with the latest accessibility guidelines - being what they term ‘SENDA compliant’ and ‘Section 508 conformant’ - then this in itself is not a criterion on which institutions can differentiate between competing products.

Equally, it could be that the implications of the SENDA are not yet fully understood or embraced by FEIs and HEIs, and so accessibility is not yet seen to be an important distinguishing attribute.

Other data from this study – in particular the perceived need for increased awareness of disability and accessibility issues on the part of both teaching staff and management (see section 7.7a) - supports the latter interpretation.

It is possible that the lack of awareness about disability generally means that the issue of accessible e-learning falls between the stools of disability services, ILT services and ICT services, with nobody taking ownership, and consequently the issue not being incorporated, for example, into institutional procurement processes.

A number of initiatives have been directed at improving levels of awareness about disability and accessibility issues in education. For example, the Teachability project (University of Strathclyde 2001), the IDEAS project (University of Aberdeen 2001) and the DEMOS project (DEMOS 2003) all address issues of disability awareness in the HE environment. The University of Wales Institute Cardiff has produced an extensive good practice guide on inclusive education (Doyle and Robson 2002), and Ferl have produced a series of 13 short guides on inclusive learning and teaching (Ferl/TechDis 2003) with a focus on information and learning technologies.

Despite these and other initiatives, lack of awareness is still seen to be a major issue by our survey respondents:

[There is] a general lack of awareness about accessibility issues: a narrowness of view about what constitutes a disability (often fostered by accessibility advocates identifying closely with one form of disability); and lack of resources or strong guidance from the top to give this work priority. (f19)

I can’t think of any reason at the moment why teaching staff should be particularly aware about disability or accessibility other than when it hits them personally and individually. Nothing has come at department, school or university level to say ‘When you design a VLE page you must remember the following about accessibility’. (i03)

Most staff should be aware of accessibility as an issue, even if not the details. At the moment, not all staff consider it to be important. (c05)

There needs to be a lot more sharing of expertise across institutions and between governmental agencies to ensure that academics are aware of [disability issues]. I don't think that some academics are even aware that putting printed handouts on coloured paper can make a difference –never mind how to change the web pages! (f02)

7.3 Accessibility testing

Questions 13 and 14 asked whether the VLE and/or its content had been evaluated for accessibility. Figure 5 indicates the responses.

Figure 5 is a bar chart showing 73% of 43 institutions testing their VLE software for accessibility, and 25% testing the VLE content for accessibility

Figure 5. Percentage of institutions testing accessibility

Recent research by the Skills for Access project found that 68% of their respondents tested at least some of their e-learning resources for accessibility, although for many this was only 'when time permits' or 'when there is a disabled student' (Skills for Access 2003).

A quarter of respondents in this study said that they tested the content put into the VLE. None of the respondents tested content accessibility as a matter of course, but rather described testing as ‘ad hoc’, ‘when requested’ or ‘checking of odd samples’.

The study sample, which was addressed to specialist VLE and accessibility interest groups, is likely to be more aware and active around accessibility issues than the mainstream of post-16 education. It is possible that the incidence of VLE content testing across the whole sector is lower than the figures in this small sample suggest.

7.4 Accessibility of the VLE

Question 13 asked those 35 respondents who had tested their VLEs ‘Has your VLE proved to be accessible?’

Of the 22% who said their VLE was accessible, two verified this using automated checkers (Bobby and unspecified), two had done their own accessibility testing, including using screen readers, and four relied on certification of W3C WAI and/or Section 508 compliance by vendors.

There has been some comparison between VLE developers on the subject of accessibility. Cann and colleagues (2003) surveyed seven VLE developers on their accessibility policies, accessibility advice, and their views on adherence to relevant guidelines. They were also asked about provision within their product of a series of accessibility features relating to general usage, image maps, tables, frames, applets and multimedia.

The research suggests that US vendors (Blackboard and WebCT) ‘have made strong commitments to improving the accessibility of their products in response to the legislative requirements of Section 508’ (see section 4.2d). UK vendors, including Granada Learnwise and Fretwell Downing, ‘have also made recent strides towards accessibility according to their corporate policies, and are working on programming issues and user guidance to users’ (Cann et al. 2003).

Other research has tested the products themselves for accessibility. The market leaders Blackboard and WebCT have been the subject of most reported research (for example SNOW 2000a, Johnson and Ruppert 2001, Pearson and Koppi 2001, Evans and Sutherland 2002, Jezierski [undated]). COSE, a VLE developed at the University of Staffordshire has also been assessed for accessibility (Stiles 2001). All these studies revealed problems for disabled students accessing VLE content, problems which are examined in more detail later in this section.

In our survey, 78% of respondents said their VLE was only partially accessible or was not accessible. These respondents were asked what specific accessibility problems has been encountered. Figure 6 illustrates the range of problems cited, and the following sections address each category of problem in turn.

7.4a. Frames

The most common reason for poor accessibility was the use of frames, mentioned by a quarter of respondents.

There has been a great deal of discussion concerning frames, much of it originally sparked by usability expert Jakob Nielsen in articles such as ‘Why frames suck (most of the time)’ where he argued that frames ‘broke the unified model of the web….. because the user's view of information on the screen is now determined by a sequence of navigation actions rather than a single navigation action’ (Nielsen 1996).

In relation to screen readers, there are two issues regarding the use of frames, one relating to assistive technologies specifically, the other to general usability issues. Before 1997, versions of screen readers such as Jaws could not read web pages presented in frames because they could not determine where the frame boundaries were. Then Jaws 3.31 introduced the ‘virtual cursor’, which allowed frame beginnings and ends to be spoken. Other popular screen reader technologies also incorporated this function (Octon 2003).

However, these advances do not necessarily mean a screen reader will produce comprehensible results for the user. This still relies on web developers understanding how a screen reader will process the frames, and then constructing their code to present the contents in logical order, and always providing frame names and titles. Commenting on this issue, Norman Octon of RNIB says:

Frames need not be a problem as long as: they are in logical order; they are marked up properly; and there are not too many of them. I have come across web pages with 15 frames, which is nonsensical. (Octon 2003)

Our respondents encountered both accessibility and usability problems in relation to frames:

[The VLE] is accessible to most up-to-date screen readers, but there are problems with older assistive technologies (e.g. those that can't cope with frames). (f19)

There are some issues about the frame structure and how this will relate to screen readers. (f02)

The system has a frames-free version which addresses some issues, though this is superficial, as screen readers work better with the frames version! (f08)

Craven and Brophy recently tested digital library interfaces with blind and visually impaired students, and looked at the issue of frames:

While WAI Guidelines advise on the use of frames, it should be understood that the critical issue tends to be the complexity of pages and the logical relationship between areas on the page. Thus a page containing frames may not itself be problematic; a page containing a number of frames which require a user to make mental links between them will be inaccessible. Designers should consider the steps needed to navigate within pages since this is the most crucial determinant of accessibility. (Craven and Brophy 2003)

The SNOW project at the University of Toronto makes a similar point regarding WebCT:

With [each new version of WebCT] there has been a significant increase in the features and utilities available in the student interface. This has caused concern among screen reader users, for whom the complexity of the framed layout significantly increases the cognitive load. While it is technically ‘accessible’, it is still very challenging for an inexperienced screen reader user to become comfortable and oriented in the WebCT courseware environment. Until such time as screen readers, browsers and courseware are developed to a point of seamless integration with consistent support for access strategies, complex interfaces such as WebCT will continue to be a challenge. (SNOW 2000b)

One of the survey respondents, asked about the most pressing accessibility issue, simply said: ‘Tell the VLE developers not to use frames’ (j04).

7.4b. Synchronous communication

The principal synchronous communication functions in the virtual learning environments in this research are chat facilities and interactive whiteboards, many of which use Java applets as their underlying technology. As a group these functions were the most problematic in terms of accessibility:

The online chat and virtual classrooms are Java-based, and not accessible to screen readers. Even for the so-called accessible versions, a very high level of IT skills is needed from the users. (f20)

The synchronous nature of these tools clearly means that timeliness – speed of access, speed of response – is vital in order to use them meaningfully. Online chat is therefore not just a problem for users of screen readers, but for students who may read, assimilate, compose responses or input more slowly, either because they need to do so in an alternative medium, or simply because of their personal learning style. Technical fixes will not address all these problems, so while a function may be accessibility standards/guidelines compliant, it does not make it universally usable. As the last respondent above also points out, a very high level of IT skills is needed from users to access these functions using assistive technologies. The issue of user skills is examined in section 7.7.

Several respondents suggested realistic approaches to the use of alternatives where tools or functions are not usable by some students:

We need to acknowledge that some 'extreme' online tools - chat and whiteboard - may not be suitable for all, and alternatives must be provided. (f05)

The baseline for us at the moment is to build as much accessibility into the main content as we can. Then identify which bits are not accessible, and maybe produce a whole alternative if necessary, but within the context of your learning objectives for that particular activity. If it is just an extra thing that is not central to the progression through the course, then we need to acknowledge it and document it – for example, ‘a blind student at this university will not be able to do this activity but that will not have a major impact on them being able to pass the course’. Or to say ‘a blind person can’t do this but this is central and is not accessible so here is the alternative’. It must all be in the context of overall learning objectives. (i02)

7.4c. Usability and accessibility

Of equal importance to our survey respondents was the issue of poor usability hindering accessibility. Usability is defined at the degree of efficiency, effectiveness and satisfaction with which a user can achieve their intended task/s using a system.

Five respondents in this survey did not rate the general usability of their VLE very highly, for example:

The VLE is too complicated. This is not disability specific – the general usability just seems poor. (c01)

The web has been estimated to be three times less usable for people with visual impairments than those without (Coyne and Nielsen 2001). VLEs have been shown to be six times as difficult to use ‘due to their complexity, and the expectation that learning will take place during use’ (Evans and Sutherland 2002).

Proponents of user-centred design emphasise the need to promote good usability – sometimes termed universal design or design for all – as opposed to accessibility in isolation (see for example Frontend 2001, Pearson and Koppi 2001, Neumann 2002, Kelly and Craven 2003). While a VLE resource may be technically accessible, that does not mean it will be usable:

Our materials are making good progress towards becoming compliant to relevant standards. However, our biggest problem is not strict standards adherence, but…. ensuring good, logical usability (it goes without saying that 'minor' visual anomalies, inconsistencies and annoyances generally present a far greater problem when, for instance, accessed by screen-readers). (c01)

As Pearson and Koppi point out, the solution to this is encapsulated by the concept of learner-focused design:

Learning environments have requirements for accessibility beyond functional considerations and the use of ‘alt’ tags. Care needs to be taken with navigation, structure, content design and communication aspects, and learner-centred design is crucial to ensure that online learning is accessible to students with disabilities. (Pearson and Koppi 2001)

Stiles, when evaluating COSE, a bespoke VLE from Staffordshire University, also comments:

The success of the [VLE’s] metaphor and the ease with which learners can intuitively navigate between components and functions plays a major role in its success. To be truly accessible, the metaphor and navigation must be independent of disability. (Stiles 2001)

It must be borne in mind that VLEs are intended to help people learn. Usability problems in VLEs may merely be short-term frustrations or annoyances, but they may also be the source of cognitive overload, impinge on the whole learning process, causing ‘at least reduced quality of learning and at worst total despondency and reluctance to engage in any further learning experiences’ (Evans and Sutherland 2002).

However, one caveat was raised concerning the concept of universal design. One of its central claims is that accommodations made for specific individuals can greatly benefit all. The example of cut-down roadside curbs is often given – a design accommodation for wheelchair users that benefits people with pushchairs, shopping trolleys, bikes and so on. However, one respondent pointed out that:

A difficulty for lecturers in working with disabled students is that each student is different, and what works for one will not necessarily work for another - despite the much vaunted claim that access work benefits all students. (f06)

While good usability will certainly benefit all, the diversity of possible accessibility accommodations for individual students – and the associated time and resources involved in delivering these individual accommodations - is an important issue, which is addressed in more detail in section 7.7h.

7.4d. Navigational issues

Poor navigational design is often included under the general heading of poor usability, but navigational complexity was singled out as a particular problem in VLEs by several respondents, for example:

Screen readers have a minimum of five levels of navigation to cope with when using a VLE. (f20)

There are too many screens to navigate through to reach where you are going. (f28)

Evans and Sutherland (2002) note that navigation was a particular issue for students with visual impairments using WebCT. Stiles also pinpoint navigation as the most problematic part of VLEs, specifying the following requirements:

7.4e. Flexibility of display

Several respondents commented that aspects of the VLE’s appearance on screen were not accessible in their default settings, and sometimes not changeable because they are ‘hard-coded’ into the VLE’s core programming:

So far the only problem we have encountered is not being able to make the screen colours different. (f32)

A fundamental aspect of user-centred web design is to allow the user control over how the content displays. This is one of the great advantages of web-based media for people who have specific requirements regarding, for example, the colour or size of elements on the page. This facility is contained within all standard current browsers. It can be incorporated into the VLE as well. If it is not explicitly incorporated into the VLE, then the VLE coding must allow the user to use their personal settings to override the VLE default. This does not appear to be the case currently with all VLEs.

Giving control to the user also opens up the possibility of ‘platform independence’ – allowing the user to access the VLE on a palmtop computer, for example. There are some basic web development processes which must be followed in order to achieve this, notably the use of stylesheets (CSS) and valid XHTML. These enable all the display elements associated with a page to be kept separate from the ‘raw data’ in the page itself. This allows a different stylesheet to be attached to the same data for output in a different style or on a different platform.

There has been much discussion about the potential of XML technologies for enhanced accessibility and platform independence:

XML allows a range of flexible stylesheet transformations. It allows simple changes to font size and colour as well as the use of complex translation grammars used to translate a presentation into entirely different modalities. Because all content in an XML document is declared and labelled, authors can create content that later can be re-styled in ways that the author never imagined. (IMS 2002)

Commercial developers are aware of XML’s potential (e.g. Blackboard 2003a). However, as Franklin (2001) has pointed out, the potential to free content of all design constraints runs the risk of losing some accessibility features, and XML based technologies will only prove accessible if best practice is followed and accessibility is built into the XML schema and the document type definitions (DTDs).

This also raises the issue of open source VLEs. Some non-commercial developers have opened up access to the underlying source code of their VLE. This means that institutions using the VLE are able to make changes and adjustments to suit their own environment and needs. Problems such as lack of display versatility noted in the above proprietary products could be solved by institutions without having to go back to the supplier. More fundamentally, a whole system can be amended and made ‘bespoke’. One interviewee, whose institution has opted for open source, explained:

We are unusual as an institution because we operate right from SVQ level though to PhD, and we need a platform that supports a lot of long-distance collaboration between staff as well as between students and staff. We also have partners with specific non-English language needs, and specific needs for international working. And we are a young institution, so we need a VLE that grows as we grow. All these factors have led us to an open source solution; not building our own from the ground up, but taking an existing open source solution and then customising it to our own needs. (i05)

7.4f. Student assessment

VLEs allow a variety of methods of assessment, including the submission of assessed work online, and a variety of test and quiz formats. Online assessment can bring a number of benefits for disabled students:

However, Evans and Sutherland (2002) found the submission of work to be one of the most difficult tasks for students with visual impairments. One of our respondents (ferl06) made specific mention of the submission process being difficult for students using screen readers. Jezierski (undated) also points to submission problems, due to screen readers not recognising the ‘submit’ button.

There are issues with anonymous marking because you cannot get rid of the student’s user ID and name, which are automatically attached to the submission. If you try and remove the name then you lose it in discussion threads too, and of course they may form part of a course assessment. (i01)

While not strictly an accessibility problem, this is nonetheless a usability problem for the institution.

Even if technically accessible, quizzes can pose different problems, most especially if they are timed. Evans and Sutherland (2002) found learners using screen readers to do a quiz spent only one third of their time actually doing the task, and the remainder of the time accessing and navigating the VLE.

Some assessments, such as drag and drop, will clearly never be accessible to some disabled learners. Again, this becomes an issue of taking into account the overall learning objectives and providing suitable alternatives (Ferl/TechDis 2003, 3).

Wiles points out that there is no specific provision within SENDA to ensure accessible online assessment, and that this is potentially serious for the student, who may only get one chance at being assessed. In addition, while there are guidelines for accessibility in more traditional assessments (hand-written, examination halls), there are as yet no accessibility guidelines for online assessments (Wiles 2002).

7.4g. Asynchronous communication

Discussion groups or bulletin boards are a key part of VLEs, providing one of the main foci for group interaction. This asynchronous communication is inherently less difficult than synchronous types, as users can take their time participating, and the technology is relatively straightforward. None of the main commercial products appeared to cause specific accessibility problems in this area, but one in-house product was still developing a discussion group function accessible to screen readers (c05).

Evans and Sutherland note some difficulties with discussion boards for students using screen readers, who tested the tool in both Blackboard and WebCT and found the latter difficult to use. (They did however point out that the comparison was slightly problematic as the students were more familiar with Blackboard than WebCT prior to testing, which may have influenced their experience.)

7.4h. Online help

The online help is not written to be universally inclusive. The developers need to review and rewrite the help files to ensure all references to purely visual elements such as icons are removed and that the instructions are written to be universally understood. This would save me [an academic/curriculum support co-ordinator] from having to devise complex e-tutorials for blind users quickly, at the time of need, when difficulties are encountered. (f05)

7.5 Accessibility of VLE content

Figure 5 showed that a quarter of all respondents said that they tested their VLE content. They reported the following problems:

Authoring tools are a generic category that includes simple office suite programmes such as Word and Powerpoint, through to more sophisticated web-specific authoring tools such as Dreamweaver. They allow content creators to who are not web-experts to put together materials for the web without learning HTML. Three authoring issues were highlighted by respondents: the lack of user-friendliness of web-specific authoring tools, the inappropriate use of these tools, and the bad HTML produced by generic tools:

The [web] authoring tools that are available are not very user-friendly. As a result, people are often using Powerpoint or Word to create content, then ‘saving as HTML’ in order to put it into the VLE. This creates very inaccessible materials, which are time-consuming for an expert to make accessible. (f08)

Our initial checks showed problems with authoring tools – for example people using [Macromedia] Authorware instead of HTML for basic menus and text sections. (c03)

We get poor accessibility [in HE e-learning] because people use crude tools and save to HTML, producing bad code. People are reluctant to change to unfamiliar tools, and there are costs involved in changing. (Kelly 2003a)

Automatically generated code frequently breaks not only W3C Web Content Accessibility Guidelines, but is also often non-valid HTML. The code can also be bloated, making files larger than they need to be and increasing download time. (Sloan 2000)

While the majority of FE and HE teaching staff create their own content, there is a large market of ready-made learning materials. One respondent said that they had bought in NLN (National Learning Network) materials, assuming they would be accessible but then found that they did not provide keyboard shortcuts (f12).

Navigational complexity is already a problem cited with the structure of the VLE (see section 7.4d) but it gains another dimension when the content itself contains another set of navigational elements:

Screen readers have a minimum of five levels of navigation to cope with when using a VLE: it is difficult to manage both VLE navigation and in-built navigation in course materials themselves. So, our new course materials are allowing navigation to be done by the VLE [rather than adding an additional level]. (f20)

‘Alt tags’ are the ‘alternative text’ in HTML that should be added to images, so that when a screen reader, or any browser with the images turned off, accesses the page, a description of the image (and hopefully also its purpose if necessary) is included. They are a ‘bare minimum’ in terms of web accessibility, and one respondent commented that teaching staff did not incorporate them into their VLE content. In this instance it seems that the content creators lack the knowledge to include them of their own accord, and the authoring tools they are using do not require them to be included.

The results above are only for the content that is actually checked. Of those respondents whose VLE content was not checked, or who did not know if it was or not, it is probably safe to assume that the same problems would manifest. The lack of accessibility of course content is clearly a problem.

I check odd samples of content - I perform validation tests and get some expert usability reviews. I usually get very depressed about the results. (f19)

7.6 Content creation process

Figure 7 shows the results to question 16, which asked respondents: ‘Briefly, what is the mechanism for getting course content onto the VLE?’.

The responsibility for creating VLE course content lies overwhelmingly with teaching staff in both FE and HE. There are some possibly significant differences between FE and HE, with further education appearing to spread the responsibility a little more, by involving ILT (information and learning technology) or ICT (information and communications technology) staff, either in conjunction with teachers or in taking primary responsibility themselves.

Nonetheless the majority assumption across both sectors seems to be that content is created and uploaded by teachers, with varying amounts of specialist support.

The survey then asked whether any guidance on accessibility was provided for VLE content authors. Twenty-six (60%) of respondents did provide some accessibility guidance for authors, detailed in figure 8.

Almost half of those respondents who provided some form of guidance had produced their own guidelines (see Appendix 6 for more details on accessing some of these guidelines). These took the form of content within the VLE itself, or on the intranet, or institution website. Eight respondents said that the main form of guidance was individual advice and support from other staff in either ILT or ICT support.

Five respondents said that accessibility issues were included in either VLE training, or in more general staff development programmes. In context of the overall sample, this figure implies that about 10% of all responding institutions incorporate web accessibility issues in training for the relevant staff.

Pearson and Koppi (2001) comment: ‘There is a paucity of advice available specifically aimed at the design and development of accessible educational courseware for academic developers’.

7.7 Origins and priority of accessibility problems

Question 18 of the survey asked respondents to rate their level of agreement with seven suggested ‘reasons for the lack of accessibility in courses delivered by VLEs’. As Figure 9 shows, the average rating showed agreement to varying degrees with all the suggested reasons. Respondents were also asked if they had other reasons to put forward. The following suggestions were made:

Question 19 asked: ‘Of the problems outlined in question 18, including any additional ones you may have raised, what would you say needs most urgently addressing, and why?’ Responses are indicated in Figure 10; they include the six most strongly supported reasons offered for rating in question 18, as well as some of the additional reasons respondents were given space to add.

Figure 10. What are the most urgent problems in implementing VLE course accessibility (Several respondents cited more than one problem in answer to this question)

7.7a. Low awareness of disability/accessibility issues

As Figure 10 shows, the problem cited most often in this study as being the highest priority was a general lack of awareness in FE and HE institutions about the whole issue of disability access, and how it applied to e-learning generally and VLEs in particular. This finding was cited unprompted - it was not among the seven reasons respondents were asked to rate - by almost a quarter (23%) of all respondents. Typical comments included:

Awareness is the most urgent problem. Because without that widespread awareness of the general issues and the need to act, there will be no pressure to find resources and mechanisms to make content accessible, whether on a VLE or elsewhere. At present, most people, even if they are willing and interested, would not know where to go for advice, information or support. (f19)

Awareness is possibly the biggest issue. People may be aware of the term 'accessibility' but are not really aware of what is involved in achieving it. (f08)

I believe that raising awareness is the most important issue - making course designers and deliverers aware of what they can do to make material and courses accessible, and making students aware of what we can do with them. (f28)

Unless teaching staff are affected by the issues of accessibility, either in a personal or professional context, the need to adapt or create material to accommodate users of assistive technology is translated into an additional work burden. (f05)

Awareness of accessibility issues for content authors [is the most urgent priority], not least because of the implications of the SENDA. (j02)

This lack of awareness may seem surprising in view of the fact that SENDA has been in force since 2001, accessible learning is part of the sector’s quality audit systems, and the issue of widening FE/HE access to disabled learners has been on the educational policy agenda for nearly a decade, certainly since the 1996 Tomlinson report.

Two recent studies seem to confirm a lack of awareness with regard to university web sites – the institution’s main point of information access - which were found barely to reach basic levels of accessibility (Kelly 2002, Nomensa 2003). Stiles adds that ‘a classic problem [with VLEs] was that unless an institution pays attention to accessibility issues in the design and organisation of its web site, students may not be able to navigate to the page from which a VLE is launched’ (Stiles 2001). (In our study, nine of 43 institutions did not have a general web accessibility policy (question 9).)

It should be noted that lack of awareness about disability/accessibility is not restricted to academia in the UK. Recent research by Mediasurface (2003) shows 81% of UK businesses are still not compliant with the 1995 Disability Discrimination Act, and 54% were unaware of the DDA’s provisions on web accessibility.

Progress is being made in FE and HE. A JISC/UCISA (2003) survey found that half (51%) of all respondents across FE and HE provided ‘online systems which support a limited range of accessibility needs of students with disabilities’. But despite sector-wide efforts by bodies such as JISC (for example through TechDis) and BECTa (for example through Ferl) to raise awareness, and a number of individual research projects and initiatives - many of which are referenced in this report – this study and other larger studies point to a gap between ‘best practice’ and ‘mainstream reality’.

7.7b. Inaccessible VLEs

The study respondents attached the next highest degree of importance to inaccessibility within the VLE product itself, or a lack of support from the developers to enable users to ensure accessibility (Figure 10). Typical comments included:

VLE designers need to improve/develop products to be fully user - author and student - friendly and accessible – ASAP. (f27)

It would be better if the VLE producers made accessibility a very high priority… and made the system more user friendly for all of us. (f28)

[The most urgent problem is] VLE not being accessible; companies need to spend more time researching and developing useful interfaces. At the moment [VLE development is] product-driven and not end-user driven. (f31)

[The most urgent problem is] accessibility of the VLE. If this is not addressed then none of the other things will lead to accessible learning experiences. (c07)

The problems caused by inaccessible aspects of the VLE are often outside the control of the institutions, because they are hard-coded into the programming (which is not open to institutions to change themselves):

[The most urgent problem is] VLE designers needing to address accessibility. Although we can make accessible materials to put into it, we cannot alter the structure/code of the VLE to rectify its failings in terms of accessibility. We encourage academic staff to focus on the pedagogical aims of the online course/materials they are developing, and any usability or accessibility problems we experience with the VLE impede this process as well as making it more difficult for students (whether disabled or not) to access the online courses. (j03)

You can pay the vendor to work on accessibility problems you cannot address yourselves, or if you get enough other users of the system together you can press for them to make the changes. But we haven’t got much control – we have to rely on them to do it. (i04)

Accessibility problems hard-coded into the VLE software… are not rectifiable by the purchasers. This has a profound impact, in that these problems can render inaccessible all the content held within a VLE, regardless of whether or not that is accessible in itself. (Stiles 2001)

All the above comments make an interesting comparison with the results of the investigation by Cann and colleagues investigation into the policies and stated intentions of the main VLE developers, who all claim to be addressing accessibility problems. (Cann et al. 2003)

The mismatch between manufacturers’ claims and institutions’ experiences could be because:

This second issue was raised by two respondents in this study, in relation to VLEs themselves and the content placed on them:

[There is]… a general obsession with standards compliance in the belief that just because an automated tool says the VLE complies with something, it is therefore accessible. (c01)

My concern about guidelines is that they are all very well, put together by experts and so on – but do they actually benefit the user at the other end? I have seen developers spending a great deal of time implementing guidelines, and validating their code and so on, and then the end-result not being of any use to the student. I really worry about guidelines – that they don’t capture the real world context, how users really work. (i02)

All these issues relate back to the point respondents made about accessibility needing to go hand-in-hand with usability (see section 7.4c): unless guidelines take account of user experience, in terms of platforms, operating systems, browsers and user needs, then ‘compliance’ alone will not solve the problem.

7.7c. Authors lack technical skills

While this problem was third in priority for respondents (Figure 10), it was the most strongly supported (Figure 9) of the reasons given in question 18 of the survey for inaccessible VLEs.

Content authors do not always have the technical knowledge/awareness to ensure accessibility: accessibility is still seen as being the domain of specialist disability staff within institutions and not the responsibility of teaching staff, therefore I don't believe that many content authors are aware of the issues. (c04)

[The biggest problem is] content authors not having awareness of technical issues, or the time or resources to cater for accessibility. (f15)

Problems lie in some of the content posted by teaching staff rather than with the VLE itself. A common example is failing to use alt text for images thus making them unsuitable for screen-readers. (j02)

[The biggest problem is] teaching staff IT skills. Level of IT literacy is worse in the UK than in education in many EU countries, and a long way behind the US. Attitudes in general towards online learning are dismissive and negative. (f17)

[Another problem is] Insufficient technical skills in the majority of tutors to have the confidence to use the VLE to its full extent and insufficient skills to realise that they are not meeting the recommended guidelines. (f32)

A number of studies have pinpointed staff skills development as a critical issue in the implementation of e-learning (for example Jenkins et al. 2001, Stiles 2002, Conole 2003, Hanson 2003, Traxler 2003a):

Most discipline-based academic course developers are stretched to keep abreast of technical developments. Often, online course development is carried out in addition to their normal workload, and they have little time to learn programming skills or understand the technical terminology used in many guidelines. (Pearson and Koppi 2002)

There seems little doubt that teaching staff currently do not have the technical skills required to ensure accessibility in the web-based materials they produce. But should they be expected to? For example, Sloan argues:

For inexperienced HTML authors, it is vital that all materials are checked by an experienced web designer who is accessibility aware… Institutions should provide a mechanism for checking all resources. (Sloan 2000)

Management [need to] recognise that an in depth knowledge of [technical web skills] cannot be gleaned overnight, and possibly that all tutors do not necessarily need all of these skills. There should be someone available to help - after all, we have reprographics teams to help with paper based solutions. (f32)

A separate unit [is needed in the institution] to work with lecturers, and transform their basic material into VLE-acceptable material. The [institution] should understand and accept that not every lecturer is able to write suitable material - instead of trying to force lecturers to produce something which may be totally unsuitable, the college should investigate buying in ready-made material. (f23)

We are considering having a central service that checks VLE content. We don’t have web design expertise amongst the lecturers, and rather than attempting to educate everybody in the latest web design and usability skills, it is much easier to have a central support service that recommends how to do things and then checks the results as well. You need to be careful, however, to avoid bottlenecking the process – you need to plan what will be centrally checked, and how often. (i05)

Web accessibility training – or rather a lack of – is not just an issue for teaching staff, but for ILT and ICT staff as well. Skills for Access researched over 200 people responsible for e-learning authoring , including ILT and ICT as well as academic staff, and found:

Fifty-three per cent were self-taught in accessible design methods; only 64 (30%) had received institutional training in accessible design, and 55 (26%) had had no training at all. (Skills for Access 2003)

This issue is examined in the next section on possible solutions to accessibility problems.

7.7d. Authors lack instructional design skills

A distinction was made by respondents in this study between web accessibility skills and instructional design skills, though the two were still seen as very linked. Typical comments included:

Just because a lecturer knows how to use a VLE, does not mean they are instructional designers. (f11)

We need to avoid the ‘quick and easy’ approach, and instead think through the [content creation] process from the beginning. An approach is needed that does not simply encourage staff to upload existing handouts, but assists them in creating appropriate, accessible materials. (f14)

[There needs to be] training for lecturers and other content providers on how to write online [e-learning] materials. (c01)

Support for this view was also contained in the agreement (3.5 on a scale of 1-5, see Figure 9) with the statement that ‘the course content is not sufficiently adapted for the VLE’.

There has been fair amount of discussion about the teaching skills required for effective e-learning, some of which have been alluded to in section 5.2c. There seems to be a consensus that a variety of skills are needed; Bonk and colleagues divide these skills into the pedagogical, the social, the managerial and the technical (Bonk et al. 2001). It is very unclear, however, how this skills should be apportioned between teaching staff, ILT support and ICT support.

7.7e. Inaccessible VLE content

The kinds of problems causing a lack of accessibility of course content are detailed above. Although respondents ‘agreed quite strongly’ (3.9 on a scale of 1 to 5) that inaccessible content was a problem in VLEs (see Figure 9), it was rated lower down the scale of importance than inaccessible VLEs (see Figure 10).

One of our respondents commented: ‘the VLE being accessible will only be useful if the content being uploaded also is’ (c01), echoing Stiles:

Regardless of the issues associated with the use of VLEs as software products and their operation and navigation, unless proper regard is paid to issues affecting the content put into VLEs by tutors for access by students, all that attention to other issues by VLE manufacturers will achieve is easy access to inaccessible content. (Stiles 2001)

For the disabled student these distinctions are merely academic - accessible content is useless inside an inaccessible VLE, and an accessible VLE cannot make up for inaccessible content. But for institutions an inaccessible (proprietary) VLE is very much outside their control (see section 7.7b), whereas inaccessible content is – in theory at least – a problem institutions can address themselves.

Respondents had a range of suggestions as to how to address the problem of inaccessible content, which are examined in section 7.8.

7.7f. Insufficient user testing

Section 7.3 showed the levels of accessibility testing being conducted in FE and HE (see Figure 5), with approximately three-quarters of institutions performing some kind of test on the VLE itself, and a quarter testing the content put into the VLE.

Respondents to the survey ‘agreed quite strongly’ (3.8 on a scale of 1 to 5 – see Figure 9) that insufficient user testing was one reason for inaccessible VLE courses.

The issue of testing begs a number of broader questions about user-centred VLE development and student-centred learning which are addressed later in this section (7.8). Several respondents to this survey felt that properly administered user testing was one of the fundamentals required for the implementation of VLE accessibility, for example:

User testing informs all other issues. We need examples of successful accessible VLEs that have been tested with users to demonstrate the principles of accessible educational materials in an accessible VLE. (f29)

The end users must be involved in testing at all stages of development. Only by including their experience and feedback will content authors and developers have any chance of making materials and VLEs accessible and usable. (f20)

Stiles (2002) asked over 100 institutions about their reasons for choosing a particular VLE. Ease of use for staff was ranked second of 26 given reasons. Ease of use for students was equal last.

Attention was also drawn to the fact that VLE developers did not seem to include sufficient testing for the end-users – the students:

[VLE] companies need to spend more time researching and developing usable interfaces. At the moment they are product-driven and not end-user driven. (f31)

7.7g. Lack of management support

Lack of resources from central services/lack of time are the main problems. (f02)

Accessibility will be improved only when the lecturing staff are allowed time to develop e-learning….. Resources need to be made available to allow them to do this. (f17)

[There is a] lack of resources or strong guidance from the top to give this work priority. (f19)

No accommodation is made for teaching staff attempting to implement accessibility. You are given no more time and no more resources. (i03)

I feel concerned about never having the resources to actually start addressing accessibility properly. We are all aware of the problems, and we want our materials to be accessible, but we get so bogged down –it sounds awful, I know – but you do get bogged down with day-to-day things. And all the time you are worrying ‘there is this big issue out there but I don’t have the resources to deal with it’. (i04)

One respondent pointed out that the allocation of resources is determined by strategic priorities:

In academia as a whole there is not enough focus on good teaching. You get rewarded more for research than for teaching. So if you create an online course you get no credit for it – either from your peers, institution or the academic world as a whole. (i01)

Kelly and Craven (2003) have pointed out that, over and above technical considerations, institutions need to address the support implications for staff expected to produce accessible web-based teaching material.

This may well have funding implications, but can also start with improvements in institution’s internal communications:

IT support staff have not been very involved in the VLE, so they do not encourage students to use it; tutors lock their VLE areas so only their students can use them. Everyone does their own thing and there is no coherent, planned college style with interactivity for the students; the IT support staff have had no technical training in the VLE, so they cannot help tutors plan and put up suitable content. (f23)

Because accessible e-learning cuts across teaching departments, ICT support, ILT support and disability support, it can end up being ‘nobody’s baby’. Queen Margaret University College undertook a ‘round table’ approach to improve staff awareness and communication concerning accessibility legislation and its impact on online teaching within the institution (Peacock et al. 2002) in order to try and address this issue.

As Phipps and colleagues point out: ‘Without strategic commitment to developments such as inclusive learning… it is extremely difficult to make the kind of institutional and cultural change that is needed.’ (Phipps et al. 2002, p33)

7.7h. Insufficient course development time

Time and resources’ are often referred to in the same breath in discussions about organisational change, but time is singled out here by several respondents in relation to specific web development processes:

Time [is one of the biggest problems] People struggle to keep on top of work as it is. Making materials accessible does take more time, and is not taken into account. (f08)

More time needs to be spent on each project. Fulfilling WCAG Priority 3 adds significantly to development time, whether at the level of careful HTML, using Flash or to provide increased user accessibility. We are exceptionally lucky to have an e-learning development team. In most FE colleges, VLE content is created by enthusiasts - lecturers working in their own time with little in-depth knowledge of issues such as WAI compliance. (f12)

Although training is always useful, it does not mean content authors will have the time to put what they learn into effect. (f15)

Recent research by the Skills for Access project confirms time as a crucial factor:

Forty-two per cent of respondents cited a lack of time as the primary barrier preventing them from creating accessible e-learning, by far the most common barrier, ahead of difficulties in developing a prioritised management plan for redesign (10%) and a lack of knowledge of the needs of disabled people (10%). (Skills for Access 2003)

7.7i. Lack of technical support for students

This issue was not volunteered in respondents’ priorities, but as Figure 9 shows, respondents did register mild agreement (3.2 on a scale of 1 to 5) with the statement when presented with it. A number of studies have shown that the user’s level of skill in using technologies – including assistive technologies – is a crucial determinant in how accessible a web resource is:

Footnotes:
(Click the footnote number to return to the text)
(23) An EduServ grant made LIFT available at no cost to all FEIs and HEIs in May 2003 (http://www.chest/ac/uk/software/lift).
(24) As Stiles (2002) notes, Java per se is not a problem with accessibility if developers follow best practice as outlined in, for example, the Sun Java Accessibility standards; see http://www.sun.com/access/articles
(25) See Beacham et al. 2003
(26) Concerning web guidelines (though not specific to accessibility), Perry and colleagues examined the implementation of IMS specifications on interoperability and found: 'Certain data elements were ambiguous... this ambiguity could cause developers to interpret the specifications in completely different ways'. (Perry et al. 2002)
(27) These categories will be used for the first time in the accessibility metadata in the JORUM+ (http://www.jorum.ac.uk) learning materials repository
(28) BRITE initiative http://www.brite.ac.uk/about.htm

Return to SENDA? Implementing accessibility for disabled students in virtual learning environments in UK further and higher education