The Formation of Group Norms in Computer Mediated Communication
Postmes & Spears (2000)
Human Communication Research, 26,3,341-371
Social construction and appropriation
An example based on technological innovation
P342 citing as an example the way in which telephone for business communication was appropriated for domestic use ‘ The process whereby users collectively establish norms is sometimes referred to as social construction’ ( ie shared meanings). ‘technologies are socially constructed implies that their properties emerge ‘ and that the process of construction will be given by the pattern of interaction and a process of appropriation by individual members.
Thinking about an Analytical frame
Social construction requires reciprocation. It is a dynamic process that develops over time, therefore need an adequate time frame to study it.
However there are difficulties with providing an analytical frame for studying the process of Social construction (SC) and appropriation
Eg
Rice (1992, p32) ‘ models of social influence and SC in general fail to provide adequate guidance as to how to identify relevant source others, operationalize different mechanisms of social influence , or specify the sources of influence at different levels’
DeSanctis & Poole , 1994 ‘ Appropriation processes may be subtle and difficult to observe’
P343 Aim : Is the use of a CMC system socially structured by emergent group norms in line with social constructionist principles. Predictions are based on a social identity approach.
The Social identity approach to how group norms form
We have multiple identities. P343 ‘ The self not only encompasses one’s individual identity, but also comprises social identities associated with valued group memberships’. Tajfel (1978, p.63) defines social identity as ‘that part of an individual’s self concept which derives from …. knowledge of… membership of a social group together with the value and emotional significance attached to that membership’. At any one time a particular identity can become salient depending on the nature of the ongoing interaction.
Social identity & prototypical group members
Individuals self refer against prototypical members. But who are these prototypical members , how do they acquire their status?
Previous views
Reicher, 1987, 1994 , Reicher, Spears & Postmes, 1995) studied groups in which the prototype was not clearly defined apriori p344 In Reicher’s view ‘ the protoyype and norms of a group emerge because group members induce them from the behaviors they observe.’ ‘ However, the way in which group norms are formed is not just the result of a passive perceptual process, but one of active negotiation and contestation within the group, limited by the group’s historical and ideological continuity’ (Reicher, 1994, 1996).
This study
P344 ‘ Combining these ideas regarding the expression and construction of social identity, we posit that group norms may emerge through interaction as a function of within-group accommodation to a prototype that is inferred from ingroup communications ( any communicative act , therefore, simultaneously defining and reflecting group norms). We propose that such processes are especially likely to occur in computer-mediated group, in which certain features of the group may reinforce the normative pull of the group ( Postmes et al, 1998).
SIDE model ‘ argues that mediated groups can be very real to their members psychologiocally despite the lack of direct physical contact’…’range of studies that show that visual anonymity does not preclude normative behaviour’ in fact it may positively impact due to the fact that the group identity becomes more salient. ( n.b meta analysis is Postmes & Spear, 1998).
Postmes et al. 1998 ) – p345 ‘ in this study a style of social interaction was activated ( either prosocial or efficiency orientated) ……. Although the style was activated successfully over time only anonymous groups converged on the primed group norm; in groups made identifiable by means of portrait pictures, no norm formation was observed.
However the process of ‘sense of identity or community’ (KRO ahha does Postmes see social identity and sense of community as synonymous?) is still unclear. Previous work has usually see sense of identity as predetermined. These authors see it as a constructive process therefore SC relevant. ( note label of a group could be important). P345 ‘ For many groups extraneous or historical clues to identity are absent or uninformative. For group members, then, the properties of the group and the behaviour within it ( the formation of social norms of conduct and of social identity itself) need to be inferred from others’ and ones own actions and responses to them’
A challenge for studying the constructive aspect of norm formation in CMC is that p346 ‘ is that interaction is naturally confounded.: It is both the source of norms and the place to observe them at work’ However Owen, 1985 and to an extent previous work of the authors , has ‘argued that consistency of communication alone can be sufficient to infer group identity’ ‘Such consistencies initially define the social identity that regulates group behaviour’ i.e. the interaction serves as the dependant variable and the emergent consistencies are the result of normative influence on the group’ (KRO DVs and SC!!)
SI predicts that style of interaction will amplify or accentuate over time’ and that it is unlikely that the norms will transfer to an outgroup.
Hypotheses
Each group will develop its own consistencies
H1 Consistencies exist in groups interacting via a CMC system, such that variations of content and form of interaction styles will be larger between groups than within groups’
H2 Group norms will emerge over time such that interactions within a group will conform more and more to that which is typical of the group’s style and content of interaction. ( note SI predicts directional changes, SC is less predictable)
H3: A group’s norms will influence communication within the group, but not communication with people outside of the group. Other norms may apply in communication with people from other groups.
Method
Context: Optional states course supplementing a regular course.
P’s followed course at 1 of 24 terminals separated by booths.
Some students would meet & ? know each other as a consequnce of f-to-f study although not specifically stated . Mean age 21 years
Email ( provided for communication with tutor) was discovered and used by students to communicate with other students 73% of usage. 245 character max for messages.
Content analysis
Counts
Words/charas
Self referents
Paralinguistic ( ?emoticons)
Topic
Sociopersonal
General
Academic
task
other
Type
Request
Complaint
Reaction
Humour
Display ( statement) of emotion
Time perspective eg coffeee now
Personal revelations
Uninhibited behavior ie flaming ( note no relationship between flaming and self awareness as previously claimed’ Confirms that context and relations between participants can result in messages that can be classified as flaming by some operatlisations of the term by outside observer although accepted as norm by the group
Message form
Abbreviations
Formalises eg thank you
Slang
Pronouns
Superlatives
Shouting
Defining the groups using cluster analysis based on the number of interactions
Onlys Ps who sent or receieved more than two messages were includes
Cluster analysis gave 25 groups, from 100 individuals. 4 groups of individual Ps then excluded. Of 21 remaining groups 10 had Ps of less than 4 and an insufficient message total leaving 11 groups , 4-9 members, for the final analysis. These 11 groups sent messages mostly to each other therefore they spatially defined themselves; some also became temporally defined.
Testing H1
H1 Consistencies exist in groups interacting via a CMC system, such that variations of content and form of interaction styles will be larger between groups than within groups’
Based on content analysis outlined above there were between group differences on all items apart from the use of pronouns. This heterogenity occurred despite the fact that overall 73% messages was socio-personal, 23% general academic, 0.76% task. Ie in this voluntary task the majority of communication between peers was socio-personal – not surprising given age and gender. Demonstrates that everyday uses ( of this age and gender) is not thwarted by the medium in communicating feelings.
Testing H2 Developments in Communication Within the groups over time
H2 Group norms will emerge over time such that interactions within a group will conform more and more to that which is typical of the group’s style and content of interaction. ( note SI predicts directional changes, SC is less predictable)
Two prototypes calculated for each group
(i) content ( excluding use of pronouns which did not distinguish between groups)
(ii) form
For example, p358 ‘ Group 3 was characterised by the number of flames and the amount of affection. Effect sizes were 0.59 & 0.538 respectively. The content prototype for the messages then becomes the formula
P=(0.359xflaming + 0.538 x affection). If a message is coded as a flame and as a display of affect , it receives the score (0.359x1 + 0.583x1).
These prototypes were then regressed against messages using the sequential nature of the message as the time line.
P358 ‘ There was a small but consistent effect that messages within the group became more prototypical in content over time. The prototypical form of messages, however appears not to change over time.
i.e. changes in group messages over time are socially determined.
Protoypes are only predictive within the group. Thus, over time, that which is characteristic of the group is expressed more strongly ‘…… “That this effect occurs for message content only might indicate that superficial characteristics of messages, such as the number of question marks, are not subject to social influence in the same way as content is ( KRO but may depend on the type of form you identify, also form as culturally determined per Gress ).
Testing Hypothesis 3 Communication outside the group
H3: A group’s norms will influence communication within the group, but not communication with people outside of the group. Other norms may apply in communication with people from other groups.
P361 ‘ One could argue that these group level differences arise out of individual habituation or social learning principles suggesting computer communication norms are fairly low –level learned responses reinforced by the environment ( Bandura)’………. ‘This view would predict that a generic response will be learned, which will then be applied regardless of the social relation between the actor and the source of influence’ ‘ In contrast, the social identity approach predicts that a change of social context leads to as discontinuity in behaviour to the extent that a change of context activates different ( social) identities and different norms’.
Only outgroup that could be identified was the staff ( KRO have the opportunity to look at optional conferences in DZX222). Communications were not prototypical but that might bee expected given the difference in status and power.
Discussion
P364 ‘
Importance of content in determining and reflecting social identity and social norms’
Between group differences are consistent with a constructionist approach to norm formation .
Authors claim that this is a naturalistic approach to CMC compared to experimental method then that is true and it does have the advantage of being less selective about which messages to include than a more qualitative approach. Emergence – to what extent does this approach and say thematic analysis meet criteria for allowing ideas to emerge?
JCMC 3 (4) June 1998
Communication Processes for Virtual Organizations
Gerardine DeSanctis
The Fuqua School of Business
Duke University
Peter Monge
Annenberg School for Communication
University of Southern California
Norms of technology use
Early on, researchers thought that electronic communication would be stripped of social context cues and would tend to be task-focused rather than relational, and free-wheeling rather than socially controlled. (For example, see Kiesler 1986, Sproull and Kiesler 1986.) But more recent studies show that relationship-oriented communication can be high in electronic media settings (Walther 1992 1995) and that electronic communication is heavily influenced by surrounding social norms (Ferrara et al. 1990). Numerous studies show that electronic communication patterns are less a function of the medium per se than of the norms, practices, and social conditions surrounding media use (Abel 1990, Lea and Spears 1991, Spears and Lea 1994). Even low channel-capacity media, such as text-based electronic mail, can be used for complex communication if the organization encourages and supports it (Finholt and Sproull 1990, Markus 1994). In addition, individuals, groups and organizations develop certain electronic communication styles or practices.
Individuals are known to develop rhetorical styles of communication that can transfer to the electronic medium. Some styles, for example, reflect high concern about confidentiality whereas others do not; hence, some individuals are more prone to be concerned about confidentiality in electronic mail than others (Gotcher and Kanervo 1997). Similarly, individual differences can determine who dominates electronic discussion, and these differences can overpower media effects (Straus 1996). Groups can develop stylistic communication patterns as a function of the particular parties involved (Jarvenpaa, Rao and Huber 1988, Finholt and Sproull 1990) and the emergent dynamics of their interaction (Ellis, Rein and Jarvenpaa 1990, Poole, Holmes and DeSanctis 1990). As a result, there are extensive variances in the ways that people use electronic communication media, whether individuals or groups (Mantei 1989, DeSanctis, Poole, Dickson, and Jackson 1993, Zack and McKinney 1995).
Many studies, including those by Weisband et al. (1995) and Wilkins (1991), demonstrate that parties project personal styles, previous experiences and social norms of interpersonal interaction into electronic conversations. The potential for conflict is substantial as communication in the virtual organization takes place across organizational and social boundaries and as the cultural and professional diversity of relationships increases. Further, the dynamic nature of the virtual organization may make it difficult for electronic communication styles to "gel" or develop new, standard norms for communicating. Participants in virtual relationships may do well to make communication norms explicit in advance and to establish procedures for reconciling differences in communication practices that emerge as they do business across multiple boundaries. As business processes are redesigned, organizations will have to simultaneously find ways to preserve the beneficial norms that have been established while promoting newer ones that are more appropriate to the redesign.
6. Evolutionary effects
There is little doubt that the dynamics of electronic communication in the enterprise may be different in the long run than in the short run. New communities with potentially differing perspectives and social realities from those of today are likely to emerge as more persons read and write on computer networks. Electronic classrooms, for example, have been shown to evolve their own communication patterns that are kindred to yet substantively different from traditional classrooms (Duin 1991, Hiltz 1994). Impressions of others, impression management, and the degree and type of relational communication all change over the course of electronic interactions (Chidambaram 1996, Hollingshead, McGrath and O�Conner 1993, Walther 1995). As noted earlier, longer-term interactions tend to improve message understanding and deepen interpersonal relationships. The evolution of relationships in the context of electronic communication occurs both "in the small" of individual relationships and "in the large" of entire communities. A question that arises, then, is the viability of rapidly configurable, or disposable, structures in virtual organizations to sustain these longer term interactions. Organizations may have to figure out which aspects of communication they can disrupt and re-arrange and which they should allow to evolve over longer periods.
It may be that electronic communication products, such as conversations and documents stored in knowledge repositories, can provide stability to otherwise tenuous relationships. Perhaps communication histories from one setting can be carried into the communication future of other settings via evolving databases. The issues of volume, load, task, and message understanding are significant, particularly for complex information exchanges and problem solving. Future research might consider how electronic communication products might be used to support the evolutionary aspects of communication in dynamic networks and reconfigurable organizational settings (Monge and Contractor in press). For example, it may be that transactive memory systems, the context-specific and unique communication processes that develop within a group and guide knowledge sharing (Hollingshead 1998, Wegner, Raymond and Erber 1991), can be somehow formalized and re-applied when groups are dissembled and re-arranged.
