3
Effects of Representational Similarity on Deindividuation and Conformity to Group Norms
in Computer-Mediated Communication
Contrary to the popular notion that computer-mediated communication (CMC) will
liberate individuals from social influence, group pressure, and power differentials (e.g., Kiesler,
Siegel, & McGuire, 1984; Dubrovsky, Kiesler, & Sethna, 1991; McLeod, Baron, Marti, & Yoon,
1997; Sproull & Kiesler, 1986), recent studies have highlighted the potential of CMC to
accentuate, rather than attenuate, social influence (see Postmes, Spears, & Lea, 1998, for a
review). According to the social identity model of deindividuation effects (SIDE), interaction via
a computer network can actually heighten group salience and subsequently conformity to group
norms, because of, not despite, the scarcity of individuating information that likely interferes with
identification with group members (e.g., Lea & Spears, 1991; Spears, Lea, & Lee, 1990; Reicher,
Spears, & Postmes, 1995; Spears & Lea, 1992). According to this model, visual anonymity
obscures personal features and interpersonal differences and thereby diminishes the relative
importance of interpersonal concerns (Postmes et al., 2001). When combined with a salient group
identity, physical isolation and visual anonymity that characterize typical CMC settings are
assumed to lead individuals to depersonalize themselves and interacting others. Consequently, it
is postulated that people come to perceive themselves and others as representatives of social
groups rather than idiosyncratic individuals and thus become more susceptible to group influence
and ethnocentrism (Postmes, Spears, & Lea, 1998; Turner, 1987).
In attempts to provide empirical support for this apparent paradox, a group of researchers
have investigated the effects of anonymity on social influence in computer-mediated small group
settings (e.g., Spears et al., 1990; Postmes et al., 2001). As predicted by the SIDE model,
anonymity, when coupled with a salient common identity, enhanced conformity to experimentally
induced group norms. For example, Postmes & Spears (1997) reported that the isolated (and thus