Recent research shows that a sense of identity could be a stronger motivation than sense of control for users to customize their web portals (Marathe & Sundar, 2011). In addition to improving the functionality of the interface to suit one’s needs, customization lets us project our identity onto the interfaces and devices that we own. Sundar (2008b) asserts that individuals’ motivation to express their identity on media interfaces stems from the general human tendency of egocentricity. As a consequence, they will be able to perceive some aspect of their self in the interfaces. As Petty, Wheeler, and Bizer (2000) suggest, the real psychological appeal of a personalized message is that the message is oriented to some aspect of one’s self and implies a connection between one’s personality and message tone. In general, personalization of both messages and interface features (such as cell-phone faceplates) serve to make the user feel unique and distinct from others, thereby promoting positive attitudes toward the interface.
When it comes to evaluating the outcomes of customization, perceived relevance of the resulting content is considered a key mediator of the persuasive effect of interactive technologies (Kalyanaraman & Sundar, 2006). Whether the user performs the tailoring (customization) or the system does it based on the user’s prior behavior (personalization), the resulting content is likely to closely match the user’s interests and be perceived higher in utility (Sundar & Marathe, 2010), thereby leading to positive attitudes toward the interface as well as the content. The effectiveness of tailored health messages relies quite heavily on this mechanism. Tailored health messages have positive effects on health behaviors, including healthy dietary behaviors (e.g., Campbell et al., 1994; Oenema, Brug, Dijkstra, de Weerdt, & de Vries, 2008), physical activities (e.g., Marcus et al., 2007; Oenema et al., 2008), cancer screening (e.g., Jerant et al., 2007), and smoking cessation (e.g., Oenema et al., 2008; Strecher et al., 2008). These studies suggest that personal relevance enhances cognitive preconditions toward message processing and thereby increases message impact by providing individualized information on behavioral factors for achieving well-being goals (Hawkins, Kreuter, Resnicow, Fishbein, & Dijkstra, 2008).
Another variable that is closely associated with relevance is involvement. When the self is the source, both the process of customization and the ensuing content are of great personal interest to the user, thereby increasing his or her involvement with the interface (Kalyanaraman & Sundar, 2006; Zaichkowsky, 1985). Therefore, the user is likely to engage more deeply with the interface and its content, as predicted by ELM and other dual-process models. This kind of processing is known to result in attitudes that are strong, durable, and more predictive of behaviors (Petty & Cacioppo, 1986). However, not all users have the same appetite for involvement, especially with the interface. While power users show more positive attitudes toward the content and website when they are allowed to customize, nonpower users show more positive attitudes when the site personalizes the content for them—a complex phenomenon involving a trade-off between convenience and concern for privacy (Sundar & Marathe, 2010).
In sum, technological features that enable users to serve as sources of information can influence persuasion not only by increasing users’ agency, identity, and self-determination, but also by ensuring that the content is more relevant and involving.
User Engagement
The concept of user involvement discussed earlier is particularly important for understanding the role of interactive media in persuasion. By calling for heightened user activity, interactivity is assumed to breed greater involvement in the interaction, leading some scholars to propose that it stimulates central processing of mediated information. For example, Sundar and Kim (2005) showed that interactively rendered advertisements promoted purchase intentions by increasing product knowledge and product involvement.
Sundar, Kalyanaraman, and Brown (2003) created three versions of a political candidate’s website that were identical in content, but differed in the extent to which they permitted contingent interaction. High-interactivity had multiple layers of hyperlinks and medium-interactivity featured two layers, whereas the low-interactivity version had one scrollable page without any hyperlinks. Participants liked the candidate more and agreed more with his policy positions when the website had a medium level of interactivity versus low or high levels of interactivity.
Surprisingly, participants’ prior level of interest in politics did not interact with interactivity to influence their attitudes, implying that interactivity has the ability to make even apathetic users get involved in the content offered by the site. However, this advantage is negated when the site is very high in interactivity. Thus, to the extent interactivity calls for greater user activity without imposing too much of a navigational load, it can encourage both apathetic and interested users to process the content centrally, rather than peripherally.
When interactivity is high though, there is some evidence to suggest that frequency of Internet use (indicative perhaps of the ability of users) makes a difference to the moderating role of involvement. Liu and Shrum (2009) showed that heavy Internet users (> 7.5 hours per week) were more likely than light users to explore the full potential of high interactivity. This difference does not matter under conditions of low involvement, because interactivity is simply treated as a peripheral cue on the interface, leading directly to positive attitudes without actually engaging the affordance.
Therefore, in theoretically inferring the effects of interactivity on persuasion using dual-process models, such as ELM (Petty & Cacioppo, 1986) and HSM (Chaiken, 1987), we have to keep in mind that the construct of involvement has two different loci—prior user involvement with content and the degree of involvement generated by interactivity features on the interface. Given that interactivity is an affordance, that is, subject to user’s interpretation and use, the former would lead to the latter, but not always. The technology of interactivity can serve to directly boost the latter through a number of mechanisms.
Sundar (2007) refers to the latter as user engagement, which is said to be affected by three different species of interactivity—modality interactivity, source interactivity, and message interactivity—in theoretically distinct ways. Modality interactivity refers to the different tools available on the interface for accessing the embedded information, tools such as hyperlinks, mouse-overs, sliders, and drag and zoom features. Together, these functional features serve to enhance the mapping ability of our sensory channels, or perceptual bandwidth (Reeves & Nass, 2000), resulting in a richer mental representation of the underlying content.
Modality-interactivity features are often seen as “bells and whistles” and can lead directly to positive attitudes toward the interface and its content, as shown in a recent study comparing interaction modalities, such as slide and 3-D carousel with plain scrolling (Sundar, Xu, Bellur, Jia, Oh, & Khoo, 2010). Another study (Sundar, Bellur, Oh, Xu, & Jia, 2011) comparing different combinations of modality-interactivity tools on a website found that users’ perceptual assessment of the interface (perceived natural mapping, intuitiveness, and ease of use of the site) predicted their degree of absorption in the site, which in turn influenced user attitudes toward the website, as well as the content in it. Therefore, the persuasive effect of interactivity as a feature of the medium rests on the degree to which it enhances perceptual representation of the information.
Source interactivity influences the level of user engagement by affording greater agency to the user. A recent field experiment (Sundar, Oh, Bellur, Jia, & Kim, 2011) showed that participants who were able to change themes of a portal site and engage in active blogging through it became more absorbed in their activities on the site and showed more positive attitudes toward it than participants who were not able to cosmetically customize the site or generate new content, but only filter existing content. When users are the prime agents of the interaction, their level of engagement with the content is significantly enhanced. The primary theoretical mechanism is based on customization leading to hig
her engagement, as discussed earlier in the context of self-agency.
Message interactivity, the degree to which the system engages users in reciprocal communication, serves to emphasize the conceptualization of interactivity in the processual sense (Burgoon, Bonito, Ramirez, Dunbar, Kam, & Fischer, 2002) with the key underlying mechanism of “contingency” in message exchange (Rafaeli, 1988). Studies have shown that users tend to pay more attention (Sundar & Constantin, 2004), process information more centrally (Sundar et al., 2003), and feel more motivated to participate in online forums (Wise, Hamman, & Thorson, 2006) when the system allows them to have a threaded interaction.
In fact, higher degree of contingency has been found to mediate the relationship between message interactivity and other psychological outcomes, including user engagement. In a recent study with a movie recommendation site (Sundar, Bellur, Oh, Jia, & Kim, 2012), higher message interactivity in the form of footprints of user actions, responsive suggestions in a search box, and live-chatting with an online agent led to greater perceived contingency and engagement with the site, which ultimately created more positive attitudes toward it and higher intention to recommend the site to others. At the level of messages, mediators such as connectedness, reciprocity, responsiveness, and specificity of responses could explain some of the reasons why users demonstrate such iterative and prolonged forms of involvement and engagement with new media.
Whether conceptualized as a modality feature, source feature, or message feature, the primary role of interactivity in the interactivity effects model (Sundar, 2007) is to create greater engagement with content via mechanisms related to perceptual bandwidth, customization, and contingency respectively. These effects are moderated by user factors, such as their expertise in using the interface and prior involvement in the content of the interaction.
Realistic Alternative Realities
Media technologies can also aid persuasion by creating alternative realities for users. Constructs such as vividness, self-representation, self-presence, spatial presence, and transportation are important for the effectiveness of persuasive technologies in that they create more realistic experiences for users, thereby affecting persuasion outcomes.
Vividness
Richer modalities create higher levels of vividness, which can change users’ perception of a source and/or message. Vividness has been defined as the representational richness of a mediated environment shaped by its formal features (Steuer, 1992). The number of different senses engaged in the interaction (breadth) and the level of resolution within each of the perceptual channels (depth) together constitute the vividness of a medium. Given that individuals use all five senses in the real world, vivid representations using multiple modalities can increase the level of telepresence, or the sense of being present in the mediated environment (Lombard & Ditton, 1997; Steuer, 1992), thereby enhancing the perceived directness of the mediated experience. Jin (2010) found that when technology offers haptic stimuli with force feedback, such as terrain effects, acceleration and lateral forces in an online advertising context, it successfully induces desired perceptions of brand personality, such as “masculine” and “rugged.” As Fazio and Zanna (1981) pointed out in their seminal article on attitude formation, direct experience leads to stronger, more persistent, and more accessible attitudes than indirect experience.
Consistent with this, Coyle and Thorson (2001) found that a more vivid website (with audio and animation) is able to maintain positive attitudes toward the website even after 2 weeks. Klein (2003) found that a product website evoked greater telepresence when it had full-motion video and audio (compared to only text and still pictures), leading to stronger acceptance of claims made on the site. In the context of computer-mediated communication, Bente, Rüggenberg, Krämer, and Eschenburg (2008) found that real-time audio and video enhance emotional closeness and interpersonal trust of the interaction partner.
Self-Representation
In addition to vivid audiovisual and haptic modalities, an increasingly common modality for experiencing games and other virtual environments is through an avatar. While text allows you to read about an event, audio to hear it, and video to see it, avatars let you experience the event through a proxy. An avatar is a computer-generated visual representation of a user that can be customized to fit any desired appearance (Holzwarth, Janiszewski, & Neumann, 2006; Jin, 2009) and possess human-like characteristics, such as speech. This can have profound implications for persuasion. By offering a representation of our own selves, avatars have been known to change both our online and offline behaviors.
In communication research, the persuasive effect of avatars has been studied from the lens of behavioral confirmation theory (Snyder, Tanke, & Berscheid, 1977), which posits that mediated human interaction is guided by one’s perception of the other, with the latter’s behaviors being affected by this perception to the point of reaf-firming the former’s expectations of the latter. In any given interaction among people—comprised of a perceiver and a target—the target tends to behave in a manner that confirms the perceiver’s expectation of the target.
By extension, in a virtual environment, the image of an avatar can dictate the avatar user’s behavior in the virtual world in accordance with the user’s assumption about how the avatar is perceived by other users. Yee and Bailenson (2007) coined the term “Proteus Effect” to signify the tendency among individuals to model their online behavior after their digital self-representation. In their study, participants who were assigned attractive avatars showed greater intimacy with confederates compared to participants assigned unattractive avatars. Likewise, they tended to be more dominant when their digital representation was a tall, rather than short, avatar.
Perhaps the more important contribution of using avatars is that they enhance our ability to vicariously experience the mediated environment. Social cognitive theory (Bandura, 2001) has long documented the human tendency to enact observed action performed by a model, based on vicarious learning of the consequences of the action. In traditional media, the model is typically another human being. But, in virtual environments, avatars can serve as models. Given that they are self-representations, the vicarious experience is likely to be even stronger. Fox and Bailenson (2009) found that those who observed their avatars gaining or losing weight in accordance with their physical exercise performed significantly more exercise in the real world than those without such vicarious reinforcement.
Self-Presence
While avatars allow users to experience mediated environments through a proxy, virtual reality (VR) goes a step further by affording self-presence of the user in those environments. Self-presence is defined as the user’s mental model of their own body being present in the virtual world (Biocca, 1997). VR is by far the richest modality in terms of heightening the sense of self-presence in mediated reality, so much so that simulated behaviors in a VR setting have become efficient therapy for curing a traumatic experience from the same behavior in the real world.
A study on the treatment of driving phobias in patients following an accident showed that those who underwent a VR simulation recorded a significant reduction in travel distress, travel avoidance, and maladaptive driving strategies compared to those who were administered a game version of the treatment (Walshe, Lewis, Kim, O’Sullivan, & Wiederhold, 2003). In a study with arachnophobes, Garcia-Palacios, Hoffman, Carlin, Furness, and Botella (2002) found that 83% of the patients in the VR treatment group (holding a virtual spider with tactile feedback), showed clinically significant improvement in their disorder compared to 0% in the control group without any treatment.
Spatial Presence
A related construct is spatial presence, which can be enhanced by navigability affordances that aid user motion within a virtual environment. Wirth and colleagues (2007) proposed that gamers use available spatial cues in the mediated universe to mentally construct a spatial situation model (SSM), which serves to shift their primary ego-reference frame (PERF) from the physical world t
o the mediated one. Spatial presence can be induced by the interior design of virtual environments, but a key element is user navigation through the mediated space. Game designers are careful to plot out various navigational pathways in order to design several absorbing experiences for the user.
The very act of navigating through a game world can constitute a compelling narrative—a primary consideration for designers of serious games for health—that is designed with the persuasive intent of exposing individuals to information in a certain sequence. Balakrishnan and Sundar (2011) found that the traversibility afforded by steering motion increased spatial presence in a virtual space, whereas a guidance tool negatively affected spatial presence. They attributed this somewhat counterintuitive result to the real-world resemblance of the steering motion and perceived complexity of the guidance tool employed on a tablet PC dashboard. The study suggests that in order to promote spatial presence among users in a mediated environment, navigability tools should be carefully designed to signify easy access to individual goals.
Transportation
By encompassing and capturing the user’s full attention, richer modalities simulate the object, story, and context of persuasion as if they were occurring in the physical world. Such immersion (Murray, 1997) in the mediated world is likely to aid the seamless integration of simulated behaviors into the behavioral script in users’ minds. In other words, the resulting mental imagery of the situations portrayed in the fictional world is so vivid that they seem to directly apply to their corresponding real-world situations. Also called transportation (Green & Brock, 2000), this heightened level of immersion experienced by individuals in virtual narrative worlds can explain the persuasive effects of rich modalities, such as interactive virtual environments and video games.
The SAGE Handbook of Persuasion Page 80