Key points are not available for this paper at this time.
From scrolling and clicking to dragging, flipping, sliding, hovering, and zooming, the wide array of interaction techniques has vastly expanded the range of user actions on an interface. Each of these interaction techniques affords a distinct action. But do these techniques differ in their ability to engage users and contribute to their user experience? Furthermore, do they affect how users view the content and how much they learn from it? We address these questions via two between-subjects laboratory experiments. Study 1 (N = 128) investigated the relative effects of six on-screen interaction techniques (click-to-download, drag, mouseover, slide, zoom, and 3D carousel) on users' assessment of—as well as their engagement with—an informational website. The site for each condition was identical in content and design, except for the interaction technique used, so that we could isolate the effects of each technique on various cognitive, attitudinal and behavioral outcomes. Study 2 (N = 127) examined the relative effects of four combinations of interaction techniques (slide+click, slide+mouseover, drag+mouseover, and drag+zoom) on the same dependent variables. Data from Study 1 suggest that although the 3D carousel generates more user action, the slide is better at aiding memory. The zoom-in/out tool was the least favored, whereas the mouseover feature fostered greater engagement among power users. Findings from Study 2, which was conducted with a different content domain, replicated the positive effects of slide and negative effects of drag in influencing user experience. Path analyses, using structural equation modeling, revealed the importance of users' assessment of the interface (perceived levels of natural mapping, intuitiveness, and ease of use), which can have significant consequences for user engagement as well as resulting attitudes and behavioral outcomes. Design insights, theories, and techniques to test and capture user experience are discussed.
Sundar et al. (Fri,) studied this question.