#研究分享#【物联网会提高人们获取的信息质量,以及积极的生活态度】

#研究分享#【物联网会提高人们获取的信息质量,以及积极的生活态度】物联网和智慧家居已经开始出现在我们的生活之中,香港城市大学的研究人员通过研究发现,物联网设备往往数量很多,但每一个设备都专注于某一功能,在这个功能上的强大会增强个人使用时的社会存在感,从而使个人对待这些设备的时候带有更多的积极态度,并且从这些设备中获取更多高质量的信息。http://www.looooker.com/archives/38226

Interacting Socially with the Internet of Things (IoT): Effects of Source Attribution and Specialization in Human–IoT Interaction

1

Abstract

This study investigates the theoretical mechanisms by which the variations in source attribution (multiple sources vs. single source) and specialization (multifunctionality vs. single functionality) of Internet of Things (IoT) devices influence the quality of human–IoT interaction. Results from a between-subjects experiment (N = 100) indicate that IoT devices that elicit the sense of multiple agencies and are specialized in a single function induce greater social presence and perceived expertise, which, in turn, lead individuals to show a more positive attitude toward the devices and to ascribe greater quality to the information transmitted by them. The results also reveal that the effect of multiple source attribution is more pronounced for individuals for whom the content of the information has low personal relevance.

The dramatic advancement of information and communications technology (ICT) in recent years has begun to expand the ubiquity and applicability of the Internet by seamlessly interlinking everyday physical objects and introducing a new era of a pervasive network known as the Internet of Things (IoT). In this novel paradigm, a thing, or a smart object, is equipped with embedded sensors and Internet connectivity, and serves as a primary building block that facilitates interactions, communication, and integration with the surrounding environment to provide intelligent, useful services and to achieve common goals (Kopetz, 2011; Yan, Zhang, & Vasilakos, 2014). The thing in IoT may be a refrigerator that tracks the expiry dates of foods and autonomously sends a reorder instruction to the grocery store, or an air purifier in a living room that informs residents of outdoor pollution and humidity levels, a color-changing light bulb that blinks when it detects an intruder, or any smart objects controllable by and connected to the Internet, that have the ability to collect and share data over the network (Kopetz, 2011; Shin, 2014). In addition to executing their designated functions, these devices may interact and cooperate with each other by, for example, utilizing the embedded sensors to form IoT-based surveillance that helps to avoid domestic incidents, or integrating the collected data to control the heating, lighting, and air conditioning and optimize power consumption throughout the house.From a conceptual aspect, the independent, yet closely interconnected, nature of these things poses challenging questions regarding the psychological orientation of information sources. The traditional Sender-Message-Channel-Receiver (SMCR) model of communication (Berlo, 1960) assumes that the sender is the source of the message and the channel is the medium through which the message is delivered to the receiver. However, this chain of communication becomes rather complicated in human–IoT interaction because IoT is a network of multiple thingstransmitting information to the receiver and, theoretically, things can be perceived as technological sources rather than mere conduits of communication—especially when they exhibit anthropomorphic features and interfaces (Sundar & Nass, 2001). If such an attribution of source is possible, then should each device convey its own agency, or should all devices share a sole agency? What are the effects of single versus multiple source attributions in IoT? In this type of interlinked network where various functions are distributed among the whole set of objects (Borgia, 2014), can the things be perceived as having sufficient expertise in the information they deliver? Drawing upon the Computers Are Social Actors (CASA) paradigm and the Modality, Agency, Interactivity, and Navigability (MAIN) model of technology effects, the current study investigates these source-related psychological questions by assigning varying degrees of agency cues that elicit social responses to IoT devices and examining the process in which the multiple source layers implemented by the cues contribute to shaping more socially meaningful and persuasive IoT experiences.

IoT as Technological Source

In their explication of online sources, Sundar and Nass (2001) argue that ‘source’ is a subjective and psychological, rather than an absolute and unconditional, concept that is shaped by what the receiver believes the source to be. Therefore, technologies (e.g., computers, televisions, search engines, interface features) can be perceived as the originators of information, or as technological sources, even though ontologically they are not autonomous sources. This holds true especially when the technology conveys anthropomorphic cues such as voice, gender, expertise, interactivity, and personality because, as the CASA paradigm suggests, human interactions with computers are fundamentally social and their computer-mediated experiences are processed in quite the same way as non-mediated, real-life experiences (Nass, Steuer, & Tauber, 1994; Reeves & Nass, 1996; Nass & Moon, 2000). According to this view, the ontological medium of communication (i.e., the computer) emerges as a psychological source of information and, consequently, the medium effects become as influential, if not more so, as the message effects observed in the traditional communication literature (McLuhan, 1964).

Guided by this conceptualization, a series of CASA studies have shown that human–computer interaction does indeed follow the social rules of human–human interaction. That is, individuals automatically apply social rules and overlearned social behaviors in their interactions with computers that exhibit anthropomorphic cues, even when they are aware that they are interacting with inanimate agents (Reeves & Nass, 1996). For example, studies have demonstrated that individuals ascribe greater trustworthiness to specialist computers than to generalist computers (Koh & Sundar, 2010; Kim, 2014), concede more to computers that express feelings of anger and happiness (de Melo et al., 2011), and respond more positively to computers with a voice manifesting a personality similar to their own (Nass, Moon, & Carney, 1999).

These anthropomorphic or social cues are typically associated with easily retrievable and uniquely human social schemas, categories, and stereotypes. Individuals tend to prematurely rely on such cues when assessing the quality of the interaction and the information acquired through it without actively processing all relevant contextual features of the situation (Reeves & Nass, 1996; Nass & Moon, 2000). Therefore, human responses to computers tend to be not only social, but also mindless, which occurs as a result of reduced attention caused by the predominant reliance on the previously established social rules and categories (Langer, 1989). Just as human–human interaction is largely and mindlessly guided by a small number of salient cues that define the situation (Langer, 1989), anthropomorphic cues in computers automatically elicit strong social responses and lead individuals to overlook the fact that computers are not genuine human actors.

Recent advancements in ICT have enabled the utilization of various smart, interactive features and the production of Internet-enabled devices that are equipped with technological capabilities similar to those of conventional computers. This suggests that IoT devices also have the potential to be perceived as technological sources, rather than mere communication channels, especially when they attempt to communicate with users in a social way through their smart, anthropomorphic features. To examine this possibility, the current study assigns two social cues to IoT devices, namely source (multiple vs. single voice) and specialty cues (multifunctionality vs. single functionality), and explores how the perceptions of source induced by varying degrees of these cues influence the quality of human–IoT interaction. Given that IoT is not a single, independent technology, but rather a network consisting of multiple things that are capable of executing multiple functions, the effects of both multiple sources and the levels of functionality on interactions between the things and users merit further investigation.

Source Attribution: Voice as a Source Cue

Extending the CASA paradigm to this study, IoT devices that communicate through their own distinctive voices may be perceived more positively than those that share an identical voice, just as information transmitted from multiple speakers and endorsers is typically evaluated as more credible. This is because (a) voice is a strong source cue that elicits robust social responses and (b) one synthetic voice equals one source and thus multiple voices indicate the existence of multiple sources (Nass & Moon, 2000; Lee & Nass, 2004). An explanation for the multiple source (voice) effect from the information processing literature is that anthropomorphic cues, such as voice, evoke the social presence heuristic, or the idea that technology users do not acknowledge the artificiality of nonhuman social actors and interact with the technology as if with a social and intelligent entity rather than with an inanimate object (Biocca, 1997; Lee, 2004; Sundar, 2008). When such a heuristic is triggered, it serves as a salient, peripheral mental shortcut that predominantly determines the validity and quality of the underlying content (Harkins & Petty, 1981; Petty & Cacioppo, 1986).

In explicating the role played by the various cues present in technology, Sundar (2008) proposes the MAIN model and identifies modality (M), agency (A), interactivity (I), and navigability (N) as four classes of technological affordances that are universally present in most digital media. These affordances are manifested in the form of surface-level features and salient interfaces that cue cognitive heuristics leading to superficial assessments of mediated content. More specifically, Sundar (2008) argues that when a technology exhibits the agency affordance via anthropomorphic features, a social presence heuristic is triggered to assess the nature and content of the interaction. For example, researchers (e.g., Lee, Peng, Jin, &Yan, 2006; Kim, Park, & Sundar, 2013; Lee, Kim, Shin, & Lee, 2015) utilized complementary personalities, social roles, humanlike appearance, and autonomy to assign agency cues to computers. They found that the presence of these cues induced a strong social presence, which, in turn, resulted in more positive evaluations of the computers in terms of their intelligence, trustworthiness, social attraction, and safety. Consequently, individuals mindlesslyrely on the activated heuristic to make snap judgments of given information (generally in a positive direction) without engaging in analytic assessments of the content (Sundar, 2008; Kim & Sundar, 2016).

However, the question that still remains is whether a greater number of voice cues increases the likelihood of the social presence heuristic being triggered; or, why is social presence expected to be more pronounced when there are multiple sources rather than one source? According to the cue-cumulation effect proposed by Sundar, Knobloch-Westerwick, and Hastall (2007), the combinatory effects of multiple cues are likely to be stronger than the individual effects of a single cue, provided the cues are (a) consistent with each other, (b) expected to trigger the same heuristics, and (c) one cue does not dominate the situation. Simply put, the presence of multiple positive cues is believed to induce a stronger, additive effect on individual perceptions as compared to one positive cue. On the other hand, the primacy effect of one cue dominating the others may also be observed if the cues are expected to trigger different heuristics and individuals, as cognitive misers, are motivated to process only those cues that are necessary for the judgment at hand (Sundar, Knobloch-Westerwick, & Hastall, 2007). Between these two distinctive cue-combination patterns, this study anticipates the cue-cumulation effect; the positive effect of voice on evoking the social presence heuristic may be further enhanced in the presence of multiple, rather than one, voice cues, because the same kinds of cues are amassed here to induce the cue-cumulation effect. Given that one voice equals one source, multiple sources (multiple voice cues) are likely to have combinatory effects of the multiple cues and elicit a greater social presence of the IoT devices than would a single source.

In accordance with the literature discussed so far, this study proposes that the use of multiple voice cues is an effective means of inducing the perception of multiple sources, with a greater likelihood of activating the social presence heuristic compared to a single voice cue. If the operation of the heuristic is indeed the underlying mechanism that guides users' social responses in the IoT context, then the social presence heuristic triggered by the voice cues is expected to mediate the effect of multiple sources and lead to positive perceptions of the IoT devices and transmitted messages, in parallel with the earlier CASA studies that demonstrated the mediating role of social presence in explaining the positive effects of agency cues assigned to computers. This study tests all these possibilities with the following hypotheses:

H1: Receiving information from multiple sources, rather than from a single source, will lead to greater social presence of IoT devices.

H2(a/b): Greater social presence induced by multiple sources will, in turn, lead to (a) a more positive attitude toward the devices and (b) a more positive evaluation of the information transmitted by the devices.

The hypothesized multiple sources effects, however, are likely to be influenced by individual differences in personal relevance and involvement. In the literature on persuasive communication, motivational factors, such as the issue relevance of information, have been known to determine the mode and intensity of information processing (Petty & Cacioppo, 1979, 1986). For example, Oh and Sundar (2015) found that antismoking messages with a higher level of interactivity induced by multiple hyperlinks elicited greater elaboration, which was then translated into more positive attitudes toward the message, but only for those individuals with low involvement in the topic of smoking. Consistent with the MAIN model framework, their finding suggests that message evaluation is significantly influenced by easily recognizable cues (interactivity cues in this case), and this cognitive reliance on the presence of relevant heuristic cues is said to be more predominant for individuals with low involvement than for those with high involvement. Similarly, when their involvement with the issue is low, individuals tend to rely on the number of people who endorse an issue as a simple acceptance cue (Petty & Cacioppo, 1979, 1986), suggesting that the multiple voices utilized to attribute multiple sources in this study may also serve as acceptance cues of the messages transmitted from IoT devices. By extension, the multiple source effect on information quality predicted earlier is likely to be more pronounced for individuals whose personal relevance to the given information is lower than those for whom it has high relevance. This moderating role of issue involvement is examined by the next hypothesis.

全文链接:http://onlinelibrary.wiley.com/doi/10.1111/jcc4.12177/full

来源:Journal of Computer-Mediated Communication

作者:Ki Joon Kim


Comments are closed.



无觅相关文章插件