|
Article Excerpt
INTRODUCTION
During the past decade, the Internet has become one of the most important computer applications for older adults (Czaja & Lee, 2001). In the year 2000, adults over the age of 60 represented the fastest growing group of Internet users in Canada (Silver, 2001). During that year, approximately 13% of all Canadians over the age of 65 used the Internet (Dryburgh, 2001), and almost one half of those (47%) used it daily. Similar growth has occurred in the United States (Teel, 1995).
One of the more common uses of the Internet is navigation of the World Wide Web (WWW), an Internet facility that links documents (i.e., Web pages) remotely and locally in hypertext forms (Freedman, 1999; Nielsen, 2000). Compared with younger users, older adults experience more frequent problems using the WWW, including finding broken links, viewing graphics, finding new information, and revisiting pages or sites (Graphic, Visualization, and Usability Center [GVU], 1998). Thus, although older adults are generally willing and able to use computers, they often have more problems than do their younger counterparts (Czaja & Lee, 2001).
This study evaluated the factors that influence navigation of a WWW site (i.e., a Web site). We chose to focus on Web navigation for three reasons. First, data have shown that the WWW is the most frequently used Internet facility (GVU, 1998). Second, difficulty with navigation is one of the largest problems for WWW and hypertext users and is characterized by problems searching for information and getting lost while navigating (Conklin, 1987; Hammond & Allinson, 1989; Kim & Hirtle, 1995; Nielsen, 1990; Vora & Helander, 1997). Third, older adults consistently demonstrate poorer performance on most computer tasks (Czaja & Lee, 2001; Kelley & Charness, 1995) and thus may be more susceptible to navigation problems on the WWW.
Most of the current research on Web navigation is based on hypertext, which differs from modern Web pages because it is primarily text based, whereas most Web pages are a combination of text and graphics, with allowances for audio, video, and animation (Freedman, 1999). Several usability problems associated with hypertext (Conklin, 1987; Hammond & Allinson, 1989; Kim & Hirtle, 1995; Nielsen, 1990; Vora & Helander, 1997) emerge because screens of text look homogeneous and generally do not provide the typographical cues of printed matter. Visual access is restricted by screen space and resolution, and linking between pages can be very complex. These characteristics of hypertext make it difficult for users to recognize information and get a sense of location in the hypertext network (Edwards & Hardman, 1989; Foss, 1989; Nielsen, 1990).
Modern Web pages can use color and graphics to provide location cues in relation to the entire site. However, these enhancements may not provide sufficient navigation support needed by older adults. Mead, Spaulding, Sit, Meyer, and Walker (1997) suggested the WWW is particularly challenging for older users because of the complexity of interactions and number of pages to navigate. Older adults show poorer performance as the number of links increases (Lin, 2001). Because Web sites are becoming larger and more complex (Nielsen, 2000), older adults may be at a disadvantage relative to the young.
One solution intended to minimize disorientation and to help users cope with increased Web site complexity is the use of navigation aids (Foss, 1989; Hammond & Allinson, 1989, McDonald & Stevenson, 1998b; Pak, 2001). However, research has shown that navigation aids are not equally effective for all users under all task conditions. Pak (2001) found that a spatial map and a step-by-step navigation aid had different impacts on older and younger hypertext users. The performance of younger adults was independent of the aids provided, but error rates were lower for older adults using the map-based navigation aid.
One reason navigation aids may not be universally effective is that they have been designed without a clear understanding of the factors that contribute to disorientation on the WWW. Not only are task conditions (e.g., Hammond & Allinson, 1989) and characteristics of the hypertext environment important (e.g., Lin, 2001; McDonald & Stevenson, 1996), but individual differences of the users will also contribute to variability in performance. This may be especially important for older adults because they represent a unique user group with special needs and characteristics (Hawthorn, 2000). Unfortunately, a limited number of studies have examined the factors that moderate age effects on hypertext or WWW navigation. Therefore, the remainder of this review will draw on research from other computer tasks as well.
Spatial Ability
Spatial ability refers to the capacity to acquire, manipulate, and use information presented in two- and three-dimensional space. Spatial abilities may help people navigate natural (Georgemiller & Hassan, 1986) and virtual environments by aiding in the creation of cognitive maps, the internal representations of the physical environment that are used for navigation (Tolman, 1948). Spatial abilities are important for mediating the effects of age on computing skill (Kelley & Charness, 1995) but may also be important for younger computer users. Garfein, Schaie, and Willis (1988) found that the best predictors for late-middle aged (49-57 years) and older adults' (58-67 years) ability to learn a spreadsheet program were spatial orientation and inductive reasoning. Pak (2001) found that young participants who were high in spatial abilities outperformed their low-ability counterparts when navigating hypertext. Vicente, Hayes, and Williges (1987) found that spatial ability was the best predictor of performance for younger participants when navigating a hierarchical file system. Spatial abilities have been shown to decline with age (see Salthouse, 1992, for a review), and this might account for some of the age-related differences on Web navigation performance.
Reading Abilities
Reading ability reliably predicts performance on computer tasks (Kelley & Charness, 1995). For example, Gomez, Egan, Wheeler, Sharma, and Gruchacz (1983) examined the influence of age, spatial memory, and reading ability on performance when using a full-screen text editor. They found that age did not predict performance and that reading ability was the most consistent predictor of all performance measures. Vicente et al. (1987) reported that after spatial abilities, reading ability was the second-best predictor of performance when navigating a hierarchical file system. Czaja, Shark, Ownby, Roth, and Nair (2001) found that verbal speed and verbal fluency were equally predictive of performance for younger and older participants engaged in a complex information search and retrieval task.
The lack of an interaction between age and reading abilities in the Gomez et al. (1983) and Czaja et al. (2001) studies suggests that reading ability may not account for a large portion of the age-related variance in performance. However, because the WWW is composed primarily of text (Nielsen, 2000), efficient readers should be able to scan text quicker, find information of interest faster, and make more effective navigation decisions. This should result in faster navigation times for better readers of all ages.
Experience and Subject Matter Knowledge
Czaja and Lee (2001) defined system experience as knowledge and skills related to the search system and interface. For the purposes of this study, this knowledge can be separated into general computer experience (e.g., how to use a mouse) and knowledge specific to the WWW. Research has shown that general computer experience can be a significant predictor of performance. Czaja and Sharit (1993) examined whether age differences exist in the performance of data entry, file modification, and inventory management tasks. Results showed that computer experience was the best predictor of several performance measures. When prior experience was controlled, the effect of age on performance was reduced significantly. The authors concluded that the amount of prior experience with computers is a far more important variable when predicting performance than is the user's age.
Another type of knowledge thought to be important is domain or subject matter knowledge. Czaja and Lee (2001) defined this type of experience as the skills, knowledge, and expertise related to a specific content domain. Greater subject matter knowledge allows users to solve problems faster and access information more quickly. Because most hypertext and...
|
