2006 Conference General Sessions

The Effect of Accessible Web Design on Able-bodied Users

Presenter #1
Denis Anson, MS, OTR
Director of Research and Development
Assistive Technology Research Institute
College Misericordia
301 Lake St.
Dallas, PA 18612
Phone: 570-674-6413
email: danson@misericordia.edu
fax: 570-674-8054

Accessible websites are shown to be as easy to use as conventional websites under ideal conditions,
• and easier to use on other forms of browser.
The World Wide Web is not only a pervasive source of information in modern society, it is becoming the preferred means of delivery for many services. An individual who does not have the ability to access the World Wide Web may face substantial delays and barriers in obtaining both public and private services.  Surveys of web accessibility have shown that only one third of tested websites provide even basic accessibility features (Jackson-Sanborn, Odess-Harnish, & Warren, 2001). While the cost of accessible design and the limitations to creative expression are often cited as reasons for not designing web pages for accessibility, the lack of benefit to “typical” users is among the most often cited reasons (Heilman, 2005).
This study tested the assumption that a website designed to comply with accessibility requirements will be as usable as a conventional site when using the standard graphical browser, but will be more usable for able-bodied individuals when using a small format browser, such as might be found in a cell- phone or PDA, or when using voiced browser, such as might be used over the telephone.

Methods
A conventional web-site intended to provide information to the members, the press, and the general
• public was selected for use in this study. The web-site was replicated following the W3C guidelines for accessible design while maintaining the look and feel of the site. A set of specific questions was created which could be answered from the web-site. 120 able-bodied subjects were recruited and randomly assigned to use either the conventional web-site or the accessible web-site via the conventional browser, a small-format browser, or a text-only browser. The subjects were asked to answer as many of the 24 questions as they could, as quickly and as accurately possible. The elapsed time and number of correct responses were recorded. They were then given a questionnaire asking them about features of the website that supported or hindered their use of the site, and about the ease of use, pleasure of use, and likelihood of revisiting the site.
Results:
Using the conventional graphical browser, visitors spent slightly longer on the conventional site (p> .05), answered the same number of questions correctly (p> .05) and took slightly longer for each correct response (p> .05) than users of the accessible site. In addition, the sites were rated equal in ease of use (p> .05), in pleasant to use(p> .05, and in the likelihood of being reused by the visitor (p> .05).
These findings indicate that an accessible website can, with attention to the look and feel, fully replicate the functionality of a conventionally-designed website for users of the full-screen graphical
small format graphical browser, visitors spent significantly more time on the accessible site (p < .05), but answered significantly more questions correctly (p < .05) and were able to get each correct response significantly more quickly (p < .05) than users of the accessible site. In addition, the accessible site was considered to be significantly easier to use (p < .05). However, while the ratings for pleasant to use (p> .05) and the likelihood of being reused by the visitor (p> .05) both favored the accessible site, they did not rise to the level of significance.
From the comments made in the experience questionnaire, it appears that the small-format browser experience was strongly affected by the presence of a large monitor. Half of the subjects in this study made explicit comments about their frustration with not being able to expand the iize of the browser window. If the physical dimensions of the screen were fully occupied by the browser, this negative reaction might have been lessened. Alternatively, it is possible that small screens are not really suitable for web browsing.
Using a text-only browser, visitors spent significantly less time on the conventional site (p < .05), were able to answer significantly fewer questions successfully (p < .05) and took substantially longer for each correct response (p < .05) than users of the accessible site. While users of neither site enjoyed using the text-only browser, the accessible site was rated significantly higher in ease of use (p < .05), as more pleasant to use(p < .05), and in the likelihood of being reused by the visitor (p < .05).
As with the small format browser, it appears that users were strongly affected by comparisons of the browser to the more familiar graphical browser. The ease of navigating with a mouse compared with use of the keyboard arrow keys was commented on by 27 of 40 respondents. In this case, we were using the Lynx browser to provide an experience similar to what would be obtained using a voiced browser over the telephone, and the use of the computer as a means of access may have affected the responses. The user of a telephone to access a website will not expect to have all of the amenities provided by a full graphical browser, so will not be as likely to complain about the browser experience. In spite of this effect, the accessible site was clearly more usable than the conventional site for users of the text-only browser.
The conventional site was strongly affected, in terms of usability, by the method used to access it. Because the text-only browser did not provide access to the navigation features, the conventional site
• was essentially unusable.
Both the conventional and the small-format browser offered full access to the features of the site. In spite of this, the conventional site was significantly less usable for able-bodied users of a small-format browser than for users of the conventional browser. While users spent the same amount of time on the site (p> .05), they were able to locate fewer answers (p < .05), and spent more time per correct response (p < .05). This difference was reflected in poorer ratings for ease of use (p < .05), pleasure of use (p < .05), and likelihood of reusing the site to locate similar information (p <05).
Users of the conventional and small format browsers had a very different experience on the accessible site. Users of the accessible site took significantly longer to locate answers when using the small format browser (p < .05), but had equivalent numbers of correct responses (p> .05), and similar times per correct response (p> .05). In addition, their ratings for ease of use (p> .05), pleasure of use (p> .05) and likelihood of using the site to find similar information (p> .05) were not significantly different.
An interesting finding is that users of the conventional website were more critical of features than users of the accessible site. For example, when using the text-only browser, the font size and colors were exactly the same between the two sites. In spite of this, all users of the conventional site said that the font was too small and difficult to read, while only half of the users of the accessible site found the font difficult, and roughly half said that the font was easy to read. It appears that the frustration of using the conventional site resulted in a more negative interpretation of all of the features. It is likely that this same negative spill-over would be directed to the organization owning the site.

 
Conclusions
The findings of the study indicate that designing websites for accessibility makes them more usable for able-bodied users in the conventional browser, in small format browsers, and in text-only browsers.
These finding are an important part of the case for accessible design. Many authors on accessible
• design, including Christian 1-Ieilmann (2005) and Joe Clark (2003) have supported the notion that accessible design has little impact on the able-bodied user of a website. The results of this study clearly show that accessible design, even when the Look and feel of the site is changed only minimally, have a significant effect on the able-bodied user, not only when using a sub-optimal browser (text only or small format), but even in the conventional graphical browser.
These results suggest that accessible design should not, in fact, be considered as an aspect of accommodation for individuals with disabilities, but as an implementation of universal design. As with other implementations of universal design, the primary beneficiaries are the able-bodied users of the device, when working in a sub-optimal environment. As an added benefit, universal design accommodates the needs of those with functional imitations, brought about by disability or by the processes of normal aging.
Since this increased usability comes at minimal cost, this study provides strong support for businesses creating accessible websites. As able-bodied users increase their demands access to the web through their cell phones, or through web-reader services, the importance of designing sites for these sub- optimal delivery channels increases, and accessible design provides guidelines for such methods of delivery of web content.

 
Bibliography
Clark, J. (2003).
Building accessible websites. Indianapolis: New Riders.
Heilman, C. (2005). 10 reasons clients don’t care about accessibility. Retrieved Sept. 20, 2005, from http://www.digital-web.comlarticles/teneasons clients dont care about accessibility! Jackson-Sanborn, E., Odess-Hamish, K., & Warren, N. (2001). Website accessiblity: A study of ada compliance. Retrieved May 25, 2004, from
http://www.ils.unc.edu/ils/researchlreports/accessibility.pdf


Go to previous article
Go to next article
Return to 2006 Table of Contents


Reprinted with author(s) permission. Author(s) retain copyright