The effect of search engine results page presentation style on user satisfaction and eye movements
Eye movement on a web page is determined by the layout of the page presented to the user (Nielsen, 2006; Shrestha et al., 2007, 2008). A typical search engine results page layout suffers from a hierarchical bias with higher fixations on the top and lower fixations on the bottom areas of interest (Granka et al., 2004; Joachims et al., 2005; Dumais et al., 2010; Curtell et al., 2007; Guan et al., 2007). This study compared the usability and eye movement data of participants searching a traditional SERP layout and two alternative layouts, grid and tabular. Results showed that while participants indicated a higher preference for the traditional layout, no differences in success, satisfaction, mental workload, or perceived difficulty were found across the 3 layouts for either informational or navigational tasks. Participants were found to look at the page more times and for longer when completing informational tasks compared to navigational tasks regardless of the layout presented to them. Participants looked at the top portion of the page more than the bottom when presented with the traditional or the tabular layout in both the task conditions. For the informational task condition, the top portion of the page was viewed for longer and for the navigational task condition, the bottom portion of the page was view for longer for both the traditional and the tabular layouts. When presented with the grid layout, participants were found to view the content vertically by column starting on the left column, then to the right: the left column was viewed more than the right column but the right column was viewed for longer than the left column regardless of the task condition. Moreover, it was found that participants fixated on the right column of the grid layout more than twice as much as they fixated on the bottom portion of the traditional layout. Compared to the bottom portion of the tabular layout, participants fixated 50% more on the right column of the grid layout. This, along with its scroll-less interface, shows the advantage of the grid layout over both the traditional and the tabular layouts. In terms of layout preference, participants equally preferred the traditional and the grid layout. It was found that the tabular layout was the easiest for the participants to parse out what element of the search results they wanted to fixate to effectively complete the given task: this meant that participants could fixate on the title and the text snippet portion of the search result and ignore the URL portion while completing informational tasks and in the same manner participants could fixate on the URL and the title and ignore the text snippet more effectively than on the traditional or the grid layouts when completing navigational tasks. This shows the advantage of the tabular layout when the user is interested in a specific type of information. The advantage of the uniqueness of their layouts may have potential with their application in proper contexts.
Thesis (Ph.D.)--Wichita State University, College of Liberal Arts and Sciences, Dept. of Psychology