Optimising website design for people with learning disabilities using 'trade-off' analysis

Introduction
Much literature from academia (e.g. Tarleton, 2004), learning disabilities organisations (Proud, 2008) and government (DH, 2005), emphasises the active involvement in society of people with Learning Disabilities (LD). These aspirations can only be achieved by the provision of accessible and relevant information - now increasingly available in electronic form.

Although the electronic medium, and in particular, the Internet, may greatly assist information provision (e.g. Adam and Tatnall, 2008), much material is out of reach of people with learning disabilities. This includes even 'easy-read' information, because of difficulties in navigating the Internet and accessing the information. Despite this problem, there little evidence regarding what design features aid website use, and advice on the subject is conflicting. Nielson, (1996), for example, recommends pages short enough to avoid the need to scroll. By contrast, Keates et al (2007) and others, urge the use of images. Pages containing such content, however, tend to be longer and therefore require scrolling. Thus one is faced with a 'trade-off' between the two attributes.

This leads to the question: how can electronically delivered information be optimally presented and organised for people with learning disabilities? This is what I seek to address in this project.

Methodology
Participants
The study will focus on those having 'moderate' learning disabilities. Such a cohort is described in educational circles as being at the 'Entry' level (Moser, 1999) People functioning at this level can:
-Read and understand simple text
-Follow simple instructions
-Get the main idea from a simple graphic (e.g safety signs).

Area of focus
This will be on access to information, rather than communication or creating content, and takes the viewpoint that information aids inclusion (Tarleton, 2004). Importantly, content understanding will not be formally measured - only 'accessible' and 'comprehensible' information will be included on the interfaces, as the layouts and presentation are the subject of examination.

Method
This will involve website usability testing with participants using and comparing accessible websites, and focusing on different attributes ('menu' lists, layout and navigation). Tasks will involve only one action, and are thus suitable for undertaking by people with low literacy skills. The tasks will be undertaken on accessible websites of different layouts (with/without images; vertical/horizontal menus etc.). Measures will be:
-Time on task: People with learning disabilities have short attention spans (LDAA, 2010) and may not wish to spend more than a short time finding information. Time on task will be used as a measure of task difficulty, rather than the simple measure of 'correct/incorrect' performance - tasks will be set that aim to be well within the ability range of the participants so they do not feel uncomfortable or intimidated. The measure has already been used in usability studies involving people with learning disabilities (e.g. Karreman et al, 2007) and is a standard measure in other usability studies.
-Preferences: Clearly, a resource that does not appeal will be underused. There is also a more philosophical reason why I wish to include preference data in the project. This is because 'Inclusion' encompasses the idea of self-advocacy and choice.

Analysis
Conjoint or 'trade-off' analysis will be employed. This is a package used in market research and never before exploited in this context. Conjoint will both determine which attributes (text size; menu position) of a site have the greatest impact on performance, and by how much. Thus this technique shows the optimum site design, taking into account combinations of all the attributes studied. The findings will provide a rich picture of how information for people with learning disabilities may be best facilitated electronically.

Apart from the academic beneficiaries of the research, as outlined elsewhere in this proposal, the main beneficiaries will be those with learning disabilities. The direct benefits they will accrue are:
-An optimum interface to use which will better facilitate information access and retrieval;
-Appropriate and accessible information on e.g. leisure activities; supported employment etc.;
-The prestige of working on the project as active participants rather than simply as research 'subjects'.

Indirect benefits include:
-Being more visible on the research agenda;
-Working in an environment where their tutors are more aware of the impact of different web interfaces on performance.

Others who would benefit in the longer term from the project are discussed below, along with a brief outline of how they might benefit:
-Parents and carers: These individuals should gain a better understanding of web behaviour and how web design can help promote effective usage
-Web developers: Formal web accessibility guidelines have been criticised for not taking into account learning disabilities (Seeman, 2006). In a study of how people with LD could be active participants in web design Helen Kennedy and colleagues (Kennedy et al, 2009) found that web developers had little awareness of intellectual disability, or of accessibility needs. The report authors were quick to point out that one reason for this ignorance was the lack of appropriate guidelines or leadership from W3G (The World Wide Web Consortium) and other web standards. This research may benefit web developers by highlighting the issues and providing suitable templates.
-Charities and the voluntary sector: The Rix Centre (a registered charity) based at the University of East London will directly benefit in that their web developers will be able to use the project results as noted above. Other charities, dealing on a day-to-day basis, will benefit in that they will have access to informational materials and - more importantly, as the project length precludes more than an exemplar set of these - be more aware of how to 'package' their own material online.
-People with other conditions (and their carers): One key attribute of this project is that the methods may be applied to other groups, such as elderly people or dyslexics. Both of these groups may require tailored design features (see e.g. Hanson, 2001 regarding elderly use and Al-Wabil et al, 2007 with respect to dyslexia) and so might benefit from better presented information over the longer term.