Quality perception and successful information seeking processes as objectives of web content presentation: a discussion of the F-shaped pattern based on an eye-tracking experiment.

Author:Lackus, Manuela

    The amount of information available through websites is rising more quickly than that available via traditional mass media. With this being the case, it is necessary to critically assess the quality of websites and their design to prove whether they meet user expectations, needs, and requirements. Websites are the key Internet interface, increasingly connecting users, customers, clients and a wide range of actors with all kinds of public and private institutions and, most recently, each other within the social network paradigm (although complementary interfaces such as Second Life, Twitter or mobile apps continue to prosper).

    In this context, an appropriate composition of websites (along with aesthetical and contentual issues) increasingly involves matters of usability, accessibility, and ergonomic design. Web usability advocates stress the importance of clarity, simplicity, and consistency in web design so that users can perform their desired operations efficiently and effectively (Cappell and Huang 2007; Nielsen 1994; Nielsen and Loranger 2006). Usability advocates have made anecdotal observations about the extent to which websites follow "good" or "bad" design practices (Fang and Holsapple 2007; Nielsen 1999; Spool 1999). But only a few attempts exist that address this issue from a research perspective and that verify high qualitative and successful user-centered design regimes for websites (Cappell and Huang 2007; Cyr 2008; Tarafdar and Zhang 2005, 2006). There is a general demand for metrics that help organizations to generate more effective websites, provide measures that online managers understand, and that academics can replicate and analyze (Neely 1998; Palmer 2003; Shapiro and Varian 1999).

    In the last decade, the usability and design of websites has received major attention, inspired by the human computer interaction (HCI) literature as well as web-specific usability research aiming at "user-friendly" designs and "ease of use" websites (Nielsen 2000). "Usability has typically taken an engineering approach in an attempt to identify a set of principles and common practices that will ensure usability is an outcome of system design. A website high in usability should generate a desirable perception of its use and an intention to use the site. Usability includes consistency and the ease of getting the website to do what the user intends it to do, clarity of interaction, ease of reading, arrangement of information, speed and layout. Appropriate design of user interfaces includes organization, presentation, and interactivity" (Palmer 2003). Nielson (1993) for example defined in his seminal book on web design five characteristics for a high qualitative design: (1) consistency of the interface, (2) response time, (3) mapping and metaphors, (4) interaction styles, and (5) media richness via multimedia and audiovisual elements. Recent research has aimed at identifying approaches to improve the ease of use and technology acceptance of the Internet (Palmer 2003; Schneidermann 1998; Smith 2008).

    Moreover, appropriate website design may have major impacts on a company's image and, even more importantly, the development of online customer trust (Hoffman and Novak 1997; Robins and Holmes 2007), loyalty (Gommans et al. 2001), and satisfaction (Cyr and Trevor-Smith 2002; McKinney et al. 2002; Simon 2001; Yoon 2002). Hoffman and Novack (1997), refer to "flow" as a characteristic of consumer behavior in computer-mediated environments, and imply that a well-designed website will arouse a user's sensory and cognitive curiosity. Koufaris (2002) applied the "flow theory" to online consumer behavior to examine emotional and cognitive responses when visiting an online store. Findings have proven that product involvement, web skills, value-added search mechanisms, and challenges (to perform to the best of users' abilities and "stretching" user capabilities) lead to (shopping) enjoyment, and ultimately to a desire to return to the site (Cyr and Trevor-Smith 2004). These findings suggest that the user experience is a complex combination of reactions that includes engagement as well as more concrete usability functions, although a deeper level of engagement through design features differs across cultural, demographical or gender dimensions, which still remain relatively unexplored (Kurniawan and Zaphiris 2005; Robbins and Stylianou 2002; Simon 2001; Zhao et al. 2003).

    In the following, we direct our attention to public institutions' websites using the example of universities. Since universities and colleges have started to use their websites as a means to recruit students, an effective website design is very important for higher education institutions to attract students and connect researchers worldwide (Weinstein 1997; Astani and Elhindi 2008). We aim to offer a first exploratory research that analyzes users by scanning the way they visit university websites.


    Recent years have seen web usability research increasingly analyze the scanning patterns of users. A pioneer of this research was (among others) Jakob Nielsen whose 2006 eye-tracking study investigated how people's eyes move when reading text-based websites. This research used corporate websites, e-commerce product sites, and search engine results (Nielsen and Pernice 2010). Nielsen discovered that eye movement, and specifically the parts of the pages that were looked at the most, followed a clear F-shaped pattern, although no direct connection to the perception quality could be determined. Figure 1 shows examples of these differences in scanning patterns.

    Nielsen explains the emergence of the F-shaped pattern as follows: The first bar is formed by a relatively close reading of the first section of text. The second movement takes a vertical path along the left perimeter of the page, where only the first words of a paragraph are read, after which the eyes jump to the next paragraph. The second bar of the F is formed by the user recognizing interesting key words which then lead the eyes farther along the middle of the site horizontally (in some cases all the way to the right side of the page). Nielsen then identifies another final movement along the left perimeter to the bottom of the website.

    There has been repeated discussion on the issue of forming a typology of eye movement in usability research. Building on Nielsen's findings, the Software Usability Research Laboratory at Wichita State University conducted further research (Shresta and Owens 2008) which clearly confirmed the F-shaped pattern with text-based websites, although not with image-based websites. It furthermore became clear that user intention does not have an influence on the scanning pattern. Regardless of whether the starting position was intentional (searching for something in particular) or non-intentional (surfing), the scanning still followed the F-shaped pattern. An important finding of this research discovers that these eye movements, as well as other activities of the human-computer interaction, are overwhelmingly unchanging, and can only be minimally influenced by "what's hot" in current web design.


    Adhering to recognized design rationales is the daily business of media designers. One of these is the orientation towards the F-shaped pattern. Although the following recognizes the F-shape itself as typical of scanning patterns, it questions the orientation of web design to this pattern. It presents the hypothesis that web content can be created by designers and perceived by users just as qualitatively or optimally even if it isn't aligned with the F-shaped pattern. This relates to the perception of attributes, which usually is subjective and differs according to how these are weighted. So this aspect of quality in particular is heavily dependent on the quality perspective held by the user (Bruhn 2004). Online media fulfill different functions, i.e. information, social, political, and economic functions. Therefore we find different dimensions of quality: (1) journalistic quality, (2) technical quality, (3) quality of design and presentation, and (4) quality of interaction. One way to test quality is to implement certain test methods such as eye-tracking technologies, data logging, self-recording, or a usability lab.

    The roots of usability testing go back to the testing of cockpit design (Fitts et al. 1950), whose aim was to investigate the connections between eye movement measurements and the cognitive activity of users. Recent years have seen eye-tracking applied and proven within usability studies on web-based stimuli (Duchowski 2003; Hornoff and Halverson 2003; Hyona et al. 2003). One of the mostused methods of eye-tracking-supported web usability methods is the generation of eye movement data while the user is searching for typical Internet information (Cowen et al. 2002, Josephson and Holmes 2002). Along with recording eye movement, the option is also available to obtain subjective feedback from the test subjects regarding the usability of individual sites, as well as gather written statements from them during and after the experiment. This was the method applied in the study presented here.


    3.1 Research Design

    The discussions that have taken place to date on this topic and their lack of connection to perception quality were the motivating factors for this study. How precisely do the users of websites perceive how they are put together, and can the F-effect really be recognized? Or, can a variation of it be seen, and if so, in what form? With the exploratory nature of this question in mind, a qualitative data acquisition method was selected. In accordance with the principles of qualitative market research, the main focus was not achieving a large sample, but getting as much out of the individual observations as possible...

To continue reading