Jeon Small, ABD, Pamela Schallau, MS, Karen Brown, MEd, Daria Ettinger, BSN, Sue Blanchard, BS, Gloria Krahn, PhD, Richard Appleyard, PhD

ABSTRACT

People with developmental cognitive disabilities (DCD) are increasingly using the Web for recreational, social, vocational and consumer needs but a reliable and accurate user model of how DCD moderate user access is lacking. The purpose of this pilot study is twofold: First, investigators present a conceptual model of several fundamental elements required for successful Web navigation. Second, we conducted usability evaluations with participants who had mild to moderate levels of DCD to examine the overall conceptual “fit” of the model. Results: The usability evaluation demonstrated that the majority of participants with DCD were unable to successfully use the W3C accessibility compliant Web site and different Web navigation problems were identified. It is clear from this study that DCD user profiles present a complex mix of user strengths and weaknesses and current guidelines do not sufficiently address accessibility for people with cognitive disabilities.

Keywords

Universal (or Disability) Access, Usability Testing and Evaluation, World Wide Web and Hypermedia.

BACKGROUND

Research of how individuals with cognitive disabilities navigate Web-based media is limited. There have been a number of studies in the last twenty years that have important implications for designing Web-based navigational interface schemes for users with cognitive disabilities. Several factors have been identified [1] that influence Web navigation: Web site interface design (including the organizational structure, navigational aids, content and terminology of the Web site), Web’s users’ expertise in using the Web, Web users’ prior knowledge of the site’s content and the Web’s users’ cognitive processing ability. It has yet to be determined how user tasks for navigating a Web environment map onto cognitive abilities or disabilities but users’ cognitive abilities are critical components to determining how electronic information can be made accessible.

Navigating Informational Structure and Hypertext environment

Navigational aids are critical in helping users traverse through information structures by compensating for errors related to a user’s query or for deficiencies that result from the system itself [2]. Informational cues that makes the users current status in the Web system transparent [3] as well as tables of contents, indices, menu dimension, horizontal tabs, hierarchical chains [4], vertical lists of links, hierarchical menus [5] and other devices based on paper-based document metaphors that affect task complexity and the knowledge structure of the Web user impact the ease of navigation.

Effective and efficient navigation requires require that users to be aware of their location in the information space, to be able to pick up a "scent" of what their next destination might be, and then to follow the right trail leading to this destination. Users need contextual information to establish a sense of location, particularly spatial context to help users decide which trail to follow next and temporal context that gives an indication of the navigation history [6].

Research Question

People with developmental cognitive disabilities (DCD) are increasingly using the Web for recreational, social, vocational and consumer needs but a reliable and accurate user model of how DCD moderate user access is lacking. Can a conceptual model of Web navigation provide researchers with a useful profile of how users with DCD navigate web sites? Similarly, what strengths and/or weakness do users with DCD bring to Web navigation landscape?

Mapping Cognitive Abilities on to Web Navigation

Figure 1 **Conceptual Model of Web Usability d** (Click Image for larger view)

Web sites continue to be built based either on artistic design parameters based on personal experience, or on usability data with limited generalization to populations with cognitive disabilities The conceptual mapping of cognitive fuctioning onto discrete aspects of Web navigtion has only been hinted at. Essentially, there are four cognitive components required for successful Web navigation [7] (See figure 1) . 1) Situation awareness - a person’s momentary knowledge of his or her surroundings. 2) Spatial awareness - a person’s awareness of how content is located in relationship to navigational devices. 3) Task-set switching - a person’s ability to quickly move from one task to another. 4) Anticipated system response - a person’s perception of how the system should respond to a user’s action.

RESEARCH DESIGN

Figure 2 -Usability Evaluation Protocol d (Click image for larger view)

Using mixed method design (e.g., think aloud and prompting) a convenient sample (n=27) of participants with mild to moderate DCD, a series of seven usability studies were held in various locations across the greater Portland, Oregon metropolitan area.. Participants were recruited from Brokerage agencies. Each participant was video taped and audio recorded. The “think out loud” sessions was started by asking the participants to think aloud during the whole session and by giving an example of how to think aloud. During the session participants were prompted to think out loud when they forgot. Sensitizing independent and dependent variables of this exploratory study are summarized in Table 1. Instrument and Procedural protocol are summarized in Figure 2.

**Table 1 - List of Sensitizing Variables Explored in the Usability Study**
Conceptual Construct
Independent variable
Moderating	Age (18-29, 30-39, 40>)
	Gender (Male, female)
	Cognitive Status (Mild, Moderate+, Moderate)
	Previous exposure to Web
Mediating Variables
Situation awareness	Use of dropdown menus	Response to navigational confirmation	Perception of location within the Web page	Perception of getting lost
Spatial awareness	Perceived clickability of Hypertext	Ability to center the mouse (Yes,No)	Lack of readability (Yes,No)	Keyboarding skill (Touch Typing, Hunt and Peck)
Task-set switching	Willingness to scroll pages	Use of BACK & HOME Key	Use of Navigation Keys	Timeliness in moving among tasks
	Knowledge of the next step
Anticipated system response	Inconsistent navigation	Non-standard interaction techniques	Willingness to read instructions	Awareness of Hypertext
Dependent Variable	Perception of Navigation Success

Table 2 - Usability Study Demographic Information and Level of Cognitive Impairment
Usability Study (n)	Computer users (n)		Total No. of Participants	Gender (n)		Level of Cognitive Impairment (Number of Participants)
	Yes	No		Men	Women	Mild	Moderate++	Moderate
1	1	0	1	0	1	1	0	0
6	3	3	6	2	4	2	3	1
7	6	1	7	5	2	3	3	1
5	5	0	5	3	2	5	0	0
8	6	2	7	3	5	1	4	3
Total=27	17	10	27	13	14	12	10	5

RESULTS - Usability Study

In Table 4 is a summary of key demographic variables used in this study. Most of the participants were in the mild (33%) to highly functioning moderate (66%) range of cognitive impairment. There was a 48%:52% split on gender; and based on self-reported data gathered in the focus groups, 78% of the participants were occasional to daily users of computers.

1. Spatial Awareness

Many participants experienced problems in finding key features of the Web sites (e.g., links to help menu, introductory links, About Us). Specifically, the ability to recognize hypertext and activate links was problematic. However, 52% of the participant could recognize hypertext links. In this group only 64% of the users could activate the hypertext link with out assistance and all knew they had arrived at the right page. The remaining 45% 1 who did not recognized hypertext links, could only select and activate the text with either step-by-step instruction (67%) or the PI executed this step. Of this group 42% recognized that they have arrived at the right page and the remaining 58% were informed that they were on the right page. With regard to this latter group, the most prevalent behavior was the random clicking of all text.

2. Situational Awareness.

Awareness about what is required is link to ones previous experience and 48% of the participants identified themselves as using Web-based media. The relationship between internal versus external links was not clear. When asked: “How do you return to the home page?” several participants closed the site when trying to return to the Home page. Two of the participants used the HOME button on the menu to return to the home page however most participants (44%) used the browser BACK button. The remaining 66% required either the PI to execute (22%) or step-by-step prompting (26%).

3. Task-set switching

The ability to task-set switch was most apparent in the participants with mild levels of cognitive impairment. This group distinguished themselves in that they: 1) All identified themselves as computer users; 2) All could locate and activate the mouse without prompting; 3) Most (56%) could type in URL with out prompting the remaining 54% could select from the drop down menu. However, Person with moderate levels of DCD moving from task to task was often labor intensive. For example, typing in and activating URLs required direct intervention for most of the participants (78%). Moving onto the browsing stage that required knowledge of hypertext, reading of text, centering and activating text, and recognizing that you have arrived on the right page to possibly start the process all over again was daunting.

4. Anticipated System Response

User anticipated interpretation of system response was almost random. Unclear navigational confirmation was particularly problematic when using the “email AddUp” feature. Most of the participants were not sure what steps they needed to take to activate (i.e., send) the message. Nor, was their any indication that the message had been sent.

Discussion

An examination of navigational patterns among persons with DCD suggests that DCD users employ a weak navigational style that is charaterized by high rate of home page revisit, high frequency of back button use and small number of page visited. Similarly, their strategies point to their overall difficulty with solving spatial problems mentally rather than in the physical world, which is turn was related to the time spent on completing a Web-based task [7]. One might hypothesize that task-set switching, particularly when working with hypertext navigation, is inherently difficult for this group because its demands of short-term memory. Finally, the weak performance on anticipated system response related tasks might be related to user experience rather than specific deficit in cognitive functioning.

ACKNOWLEDGMENTS

This study was funded by a grant from the Arc of Washington Trust Fund, as well as by a grant from the Administration on Developmental Disabilities.

REFERENCES

1. Zimmerman, D. and P. Walls. Exploring navigational patterns on the web. in Proceedings of IEEE Professional Communication Society International Professional Communication Conference and Proceedings of the 18th Annual ACM International Conference on Computer Documentation: Technology & Teamwork. 2000. Cambridge, Massachusetts: IEEE Educational Activities Department.

2. Mat-Hassan, M. and M. Levene, Can Navigational Assistance Improve Search Experience? A User Study. First Monday, 2001. 6(9).

3. Park, J. and J. Kim. Effects of contextual navigation aids on browsing diverse Web systems. in Proceedings of the SIGCHI conference on Human factors in computing systems. 2000. The Hague, The Netherlands: ACM Press New York, NY, USA.

4. Nielsen, J., Usability Engineering. 1993, San Francisco, CA,: Morgan Kaufmann Publishers Inc.,.

5. Chang, C.T., A study of hypertext document structure and individual differences: Effects on learning performance., in Master Abstracts International. 1995.

6. Utting, K. and N. Yankelovich, Context and Orientation in Hypermedia Networks. ACM Transactions on Information Systems, 1989. 7(1): p. 58-84.

7. Neerincx, M., J. Lindenberg, and S. Pemberton, Support Concepts for Web Navigation: A Cognitive Enginering Approach. ACM, 2001: p. 119-128.

Corresponding Author:

Richard J Appleyard, PhD
Oregon Health & Sciences University
707 SW Gaines Road, Suite 3213
Portland, OR. 97329-2901
Email. appleyar@ohsu.edu
Tel. 503-494-1230
Fax. 503-418-4650