Mediated social interaction refers to the interaction between two or more individuals, normally separated in time and/or space, enabled by various communication technologies. Mediated social interaction may take many different forms, depending on how many people are involved in message construction and reception (e.g., one-to-one vs one-tomany); whether participating individuals are required to be present at the time of message transfer (e.g., synchronous vs asynchronous); what kinds of modalities are being used (e.g., text vs full motion video); and so forth.
Technology-mediated communication has been with us for many years. However, it is the proliferation of personal computers and the Internet that has spawned considerable research on mediated social interaction. Not only has computer-based communication created unprecedented opportunities to form, develop, and maintain interpersonal relationships, but it has also altered the ways in which people collaborate in work groups. Thus, the following sections are devoted to computer-mediated communication (CMC), and especially how it differs from face-to-face (FtF) interaction in terms of communication processes and outcomes.
Key Characteristics Of Computer-Mediated Communication
Although recent technological advancements support a wide array of communication modalities beyond the exchange of simple texts, the most common form of CMC is still text-based. Thus, compared to FtF interaction, CMC typically lacks social context cues, ranging from nonverbal cues (e.g., facial expressions and gestures) to paraverbal cues (e.g., volume and pitch) to interpersonal cues (e.g., age, physical appearance). According to social presence theory (Walther & Parks 2002), when deprived of visual and auditory cues, people become less aware of their interaction partners and perceive them to be less sociable, warm, and friendly. In this view, reduced social presence renders CMC less effective and appropriate than FtF for socio-emotional communication.
Second, cue deficiency fosters perceived anonymity. With their identities hidden, people tend to feel freer from social constraints and become less concerned about impression management than in FtF encounters. Thus, anonymity might promote uninhibited behaviors, such as strong and inflammatory expressions (i.e., flaming). At the same time, anonymity can serve to democratize communication by liberating individuals from power differences manifested through various status cues (Joinson 1998). For example, an opinion deviate or a low-status individual might find it more comfortable publicly expressing his or her unpopular opinion anonymously, an opinion which might have been suppressed in FtF discussions.
Lastly, CMC does not require participating individuals to be co-present in the immediate environment. Rather, many variants of CMC, such as email, listserv, and newsgroups, support asynchronous interactions, allowing people from remote locations to send, read, and respond to a message at their convenience. Freed from geographical and temporal constraints, the boundary of an individual’s social network has been substantially expanded.
Self-Presentation And Person Perception
Studies have reported that people engage in greater spontaneous self-disclosure in CMC. Several explanations exist for why people reveal more about themselves during mediated communication. First, anonymity reduces perceived risks in disclosing potentially embarrassing aspects of self. Thus, those who have stigmatized identities (e.g., nonmainstream sexual preferences or political views) can find similar others more easily in chatrooms and online newsgroups and share their experience. Second, physical separation and the lack of sensory cues lead people to focus more on their inner feelings and thoughts (private self-awareness), while lowering concerns about public evaluations (public self-awareness). Supporting this view, research found that individuals’ currently possessed, but not usually expressed traits (“true self ”) became more accessible in their mind after CMC than after FtF interaction, and these traits were better acknowledged by CMC partners (Bargh & McKenna 2004).
Not only do people speak more about themselves, but they also speak better of themselves in the cue-limited environment. Although people usually present more of their ideal-self qualities to strangers than to their friends or co-workers, text-based CMC allows greater room for strategic self-presentation by eliminating a number of distractions, such as environmental stimuli, simultaneous cues from the partner, and their own physical back-channeling. Consequently, people can concentrate on message construction, making sure that their messages reflect preferred characteristics they wish to present. This selective self-presentation fosters overly positive perceptions and exaggerated interpersonal (“hyperpersonal”) expectations. Indeed, after semester-long interactions with remote partners, those who saw their partners’ photos formed less positive impressions than did those who communicated without photos (Walther & Parks 2002). In addition, not knowing that they actually interacted with the same person twice, people reported greater liking for their partner after CMC than after FtF interaction (Bargh & McKenna 2004).
One benefit of text-only interaction is the potential nullification of social stereotypes. Because these are often linked to physically salient features, such as gender and race, visual anonymity might suppress social stereotypes. Despite the absence of physical indicators, however, some social category cues seem to remain in text-only CMC. For example, mirroring gender-linked language differences observed in FtF interaction, men’s postings to newsgroups showed self-promotion, sarcasm, insults, and strong assertions, whereas women’s messages revealed supportiveness and personal orientation. Likewise, female-only discussion groups displayed more self-disclosure, personal opinions, and coalition language than did male-only or mixed-gender groups, and women relied more on graphic accents to express emotions (e.g., smiley faces) than men (Thomson & Murachver 2001). Another implicit gender identity cue is gender-specific knowledge or interest. When judging the target’s gender in anonymous CMC, people relied mostly on the conversation topic and interaction content. Moreover, even when avatars (graphical representations of a person in virtual space) were randomly assigned, in the absence of legitimate gender cues, participants inferred their partner’s gender from these arbitrary characters (Lee in press).
Once the interaction partner’s category membership is revealed, the information restrictions of the medium can amplify the category’s influence (Lee in press). For example, when the interviewee’s race was known, people rated Asian-American women as more introverted than African-American women after an email interview, although no such difference was found in telephone interviews. Apparently, to compensate for the interpersonal information filtered out in text-only communication, people turned to social stereotypes and filled in the missing information. Similarly, when gender stereotypes were activated, people exhibited stereotype-congruent behavior only when they were unable to individuate one another; when they exchanged brief biographical information with other group members, the activation of gender stereotypes had no significant effect. Taken together, these results indicate that the limited channel capacity can promote self-stereotyping, as well as category-based person perception (Postmes et al. 1999).
Group Performance And Normative Influence
In work groups, CMC has the potential to address problems common to FtF discussions: evaluation apprehension and production blocking. By allowing anonymous and simultaneous input from participating individuals, group support systems can lower concerns about negative reactions and prevent a particular member from dominating the floor. Supporting this idea, electronic brainstorming was found to be more effective for idea generation, especially when the group size was large. When group decision-making is concerned, however, a meta-analysis (Baltes et al. 2002) showed that FtF groups generally outperformed their CMC counterparts; that is, FtF groups made higher-quality decisions, took less time to reach a consensus, and yielded greater member satisfaction. Moreover, the advantages of FtF groups were more pronounced when there was a time limit and when the group size was large. Interestingly, anonymity had opposite effects on subjective and objective measures of performance, such that it decreased member satisfaction, but improved the decision quality of CMC groups.
Some studies, however, have challenged the notion that anonymity weakens normative concerns in CMC. According to the social identity model of deindividuation effects (SIDE), when there is a common group identity, the lack of individuating information (“deindividuation”) can reinforce group-oriented behaviors (Spears et al. 2001). That is, in the absence of interpersonal information that highlights individual differences within the group, people come to perceive themselves and their interaction partners as representatives of a group. Thus, they are more likely to conform to group norms and exhibit ethnocentrism. In fact, when group members were primed with certain social behaviors (i.e., efficiency vs pro-social norms), anonymous groups made prime-consistent decisions in the following discussion, whereas identifiable groups did not. Likewise, people preferred an ingroup (same university) member to an outgroup (different university) member for future collaboration only when no individuating cues were available (Tanis & Postmes 2003). Collectively, these findings suggest that normative group influence still operates in mediated interaction and can become even stronger with a salient group identity.
Directions For Future Research
To understand how CMC affects the ways in which people communicate, make sense of others, and collaborate, researchers have compared CMC with FtF interaction, treating the latter as a gold standard. An implicit assumption underlying this approach is that certain features of the communication channel (e.g., anonymity) influence individuals’ perceptions and behaviors in a rather uniform manner, representing technological determinism. However, as social information processing theory posits, people are capable of adapting to the restrictions of the medium (Walther & Parks 2002). Specifically, users have devised various graphical accents to express their feelings (emoticons) as substitutes for nonverbal cues missing in CMC. In addition, to compensate for the limited opportunities to observe others unobtrusively or to gain information about the target from a third party, CMC participants relied more upon interactive strategies than their FtF counterparts, asking more personal questions. Consequently, although it might take longer, CMC users can eventually develop as intimate interpersonal relationships as those based on FtF contacts.
Recent studies also demonstrated that different individuals use the same medium differently to different effects. For example, CMC reduced interpersonal influence by temporarily heightening the self-focus, but such an effect was found only for those who chronically experience a high level of private self-awareness. Moreover, when the task posed interpersonal risk (e.g., asking for a date), individuals with low self-esteem had a stronger preference for email over FtF communication than those with high self-esteem. Therefore, to understand better the psychological and social effects of CMC, it seems crucial to investigate how individuals’ dispositions moderate the uses and effects of the medium (Lee in press).
Due to the paucity of longitudinal studies, little is known about long-term effects of repeated engagement in CMC. For example, the SIDE model predicted and found that deindividuation amplifies the effects of group identity, but it remains unclear what will happen as people get to know each other better over time. It may be important to examine how prolonged CMC experience alters the nature of non-mediated interaction. Although some researchers have studied how online interaction influences the frequency and amount of offline interaction, it is not known how the communication procedures and conventions associated with CMC might influence the way people carry on FtF interactions, an area which awaits more research.
- Baltes, B. B., Dickson, M. W., Sherman, M. P., Bauer, C. C., & LaGanke, J. S. (2002). Computermediated communication and group decision making: A meta-analysis. Organizational Behavior and Human Decision Processes, 87, 156–179.
- Bargh, J. A., & McKenna, K. Y. A. (2004). The Internet and social life. Annual Review of Psychology, 55, 573 – 590.
- Joinson, A. (1998). Causes and implications of disinhibited behavior on the Internet. In J. Gackenbach (ed.), Psychology and the Internet: Intrapersonal, interpersonal, and transpersonal implications. San Diego, CA: Academic Press, pp. 43 – 60.
- Lee, E.-J. (in press). Wired for gender: Experientiality and gender stereotyping in computermediated communication. Media Psychology.
- Postmes, T., Spears, R., & Lea, M. (1999). Social identity, normative content, and “deindividuation” in computer-mediated groups. In N. Ellemers & R. Spears (eds.), Social identity: Context, commitment, content. Oxford: Blackwell, pp. 164 –183.
- Spears, R., Postmes, T., Lea, M., & Watt, S. E. (2001). A SIDE view of social influence. In J. P. Forgas & K. D. Williams (eds.), Social influence: Direct and indirect processes. New York: Psychology Press, pp. 331– 350.
- Tanis, M., & Postmes, T. (2003). Social cues and impressions formation in CMC. Journal of Communication, 53, 676 – 693.
- Thomson, R., & Murachver, T. (2001). Predicting gender from electronic discourse. British Journal of Social Psychology, 40, 193 –208.
- Walther, J. B., & Parks, M. R. (2002). Cues filtered out, cues filtered in: Computer-mediated communication and relationships. In M. L. Knapp & J. A. Daly (eds.), Handbook of interpersonal communication, 3rd edn. Thousand Oaks, CA: Sage, pp. 529 – 563.