Interactivity has become ubiquitous in the digital media landscape. From scrollbars in mobile texting devices to customization options in Web portals to chat functions on social-networking sites, there has been a proliferation of interactive tools affording enhanced user interaction with the system. While studies in the past have assessed the efficacy of individual tools, we have precious little generalizable knowledge about the larger concept of interactivity. How does interactivity affect user experience? Does it always ensure richer user engagement with the medium? The research will experimentally investigate three species of interactivity corresponding to the three central elements of communication -- source, medium, and message. Data will be used to articulate three ordinal levels of each type of interactivity, which will then be operationalized and used in further experiments to detect the combined effects of the three types of interactivity on user engagement as well as other outcomes of interest for both power users and regular users of Web interfaces. The research will assess the validity of conceptualizing interactivity in terms of multiple loci and more importantly, explore theoretical mechanisms by which interactivity is believed to affect user engagement.
A scientific understanding of the psychological effects of interactivity is quite critical for a society that is becoming saturated with interactive digital media. Dissemination of the proposed work will likely spawn a new wave of theoretically driven interactivity research. Research results will feed directly into design of interfaces for a variety of purposes, from learning systems to serious games. The proposed comparison between power users and regular users will shed light on interactivity's potential to bridge the digital divide. Equipment budgeted for lab studies will enhance infrastructure for research and education at Penn State's Media Effects Lab.
Interactivity is a defining feature of modern communication technologies, but there is considerable debate about its meaning and effects on users. To advance our knowledge of interactivity, this project proposed a tripartite conceptualization of interactivity by situating the concept in the three fundamental elements of communication--source, medium, and message--and testing its effects on user engagement (Figure 1). Over three years, 12 Website prototypes were created and 14 psychologica experiments, involving 1545 users, were conducted. Source interactivity is the degree to which the interface lets the user serve as the source of communication, e.g., customize one's portal or write one's blog (Figure 2). We discovered that the nature of customization makes a difference. While some functional customization to enhance task efficiency is appreciated, offering plentiful choices of tools appears to overwhelm users. However, cosmetic customization seems to unlock users' desire for self-expression. Along with active blogging, it engages users over time. Active blogging, wherein users pen original content, was found to imbue a sense of community, enhance their perceived usability of--and attitudes toward--the site. These findings not only help explain the appeal of social-media interfaces (e.g., Facebook, Twitter) that feature simple text fields, but also motivate designs for affording greater self expression to users. Enhancing opportunities for self-expression can be quite useful in online communities where under-contribution is a known problem. Our experiments suggest that, in addition to affording self-expression, interfaces should allow feedback from other users and the system. Aggregated traffic statistics (e.g., number of views, number of replies) and popularity ratings (e.g., number of stars for helpfulness) motivate user contributions to online forums. We also found that system-generated validation of a user (e.g., assigning labels like guru or veteran) encourages posting activities. In sum, source-interactivity tools boost user engagement by affording greater self-expression and providing more indicators of external validation and community. Medium (or modality) interactivity refers to the various methods of interaction offered by the interface, such as clicking, scrolling, dragging and hovering. Our studies discovered that these interaction techniques can lead to distinctive user experiences, with consequences for user absorption and learning of information and their overall attitudes toward the site. For example, the same content was recalled better when users accessed it with the slide modality rather than a zoom tool (Figure 3). Traditional techniques like slide and click were rated as being more natural and intuitive than newer ones like drag and zoom. The inclusion of 3-D carousel (whereby images rotate, as if on a carousel; see left pane in Figure 4) had both a "cue effect" (raising perceptions of interactivity, resulting in heightened expectations) and a "use effect" (increasing user explorations of the site, leading to cognitive load). We also discovered that Web analytics, while good at recording exposure, do not reliably indicate usability, thus suggesting the need for triangulating behaviors with perceptions. Our experiments show that users’ ability to process content is enhanced when delivered via modality-interactivity tools, but this comes at the cost of processing non-interactive content on the same interface. Psychophysiological data show that users have to pay conscious attention to process non-interactive content whereas they seem to process interactive content automatically. In sum, different kinds of medium-based interactivity differentially enhance user engagement by altering user perceptions of the interface, enhancing their cognitive capacity for interaction and promoting different levels of action. Message interactivity is the exchange of messages between the user and the system (human-computer interaction) or between users (computer-mediated communication). We discovered that perceived contingency determines user engagement with the interface. User attitudes were more positive, not only about their user experience but also about the information provided by the interface. One experiment demonstrated that we can successfully imbue feelings of contingency by providing visualizations of the user’s interaction history (Figure 5). This has implications for designing interfaces that approximate the contingency offered by computer-mediated communications between users. Our recent studies suggest that all three species of interactivity can have combination effects on user experience, with particular implications for persuasion and learning. One experiment focused on how interactive visualizations of data (visual graphs and charts; see Figure 6) can promote user engagement and involvement with national and international issues. Ongoing analysis suggests that our theory-based combinations of the three different types of interactivity can lead to salient learning outcomes. We continue to analyze our datasets to enhance our understanding of the psychological effects of interactivity, but the project has already yielded several theoretical insights and design ideas, resulting in the presentation of our work at major peer-reviewed scientific venues. Seven papers have been published in conference proceedings and 12 papers presented at international conferences, with six journal-article submissions, four dissertations and three book chapters. The project has provided training and education resources to several graduate and undergraduate students working in the Media Effects Research Laboratory at Penn State University.
