Virtual Neighborhoods Architecture of Online Communities

1
Virtual Neighborhoods Architecture of Online
Communities

• Reuven Aviv
• Zippy Erlich
• Gilad Ravid
• gilad_at_ravid.org
• http//www.ravid.org/gilad

2
Agenda

• Introduction
• Design, Mechanisms, Architecture
• Method
• Results

3
Design of network
mechanisms
Social Interdependence theory
Matching the predictions of network emergence
theories
Architecture
Content Analysis
Network statistical analysis of Markov models
Collaborative Knowledge construction
4
SNA viewpoints

• Global SNA
• Macro
• Cohesiveness
• Equivalence (role groups)
• Power of actors
• Range of influence
• Brokerages

• Local SNA
• Micro
• Statistical
• Dyads and triads

Aviv, R., Erlich, Z., Ravid, G., Geva, A.
(2003). Network Analysis of Knowledge
Construction in Asynchronous Learning Networks.
Journal of Asynchronous Learning Networks
5
(No Transcript)
6
The Social Capital Mechanism

• Hunt for Knowledge (social capital)
• Using efficient interactions
• E.g. bridging others
• Community works with broadcast medium
• Most efficient connection No interactions
• Passive members (Lurkers)

7
Cognitive balance
Emergence of transitive triads
cohesiveness
cliques
Creation of knowledge support
8

• The architecture of a network can be describes in
  terms of three components
• One or more relations are the fundamental glue
  between the actors
• A partition of the actors and the relations into
  2 level hierarchy of groups of actors
• A set of mechanisms shaping the relations to the
  creation of the neighborhood

9
Virtual Neighborhoods
Transitive triad
Cyclic triad
10
Method of Analysis

1. Reveal Architectural Components
2. Identify Relevant Theories
3. Identify Mechanisms

11
Method

• Analyze the recorded responsiveness data of two
  online forums of learners with different design

12
Example Two Communities

• 16 weeks each 19, 18 participants
• Parts of Open U Business Ethics Course
• Team community
• Designed for Knowledge Construction
• Tested positively by Content Analysis
• Forum Community
• Designed for support by Q A

13
Social Capital Transaction Costs

• Burt 1992, 2002
• Bridge over Holes with minimal cost
• Few single links
• linklt0
• Supported for both networks

14
Collective action

• Coleman, 1973, 1986 Marwell Oliver 1993 Fulk
  et al. 1996
• Inducements to contribute under peer pressure
• Respond to several others
• If large density centralization size then out
  star gt 0
• Supported for team network.
• Not supported for forum network because condition
  in not fulfilled

15
Exchange

• Willer Skvoretz, 1997 Hommans, 1958
• Exchange resources directly, depending on partner
  network status
• Tendency to reciprocate to resource promising
  partners
• mutuality gt 0
• Not supported for team network because there are
  no a-priory resource promising actors
• supported for Forum network because Tutor is
  a-priory resource promising actors

16
Generalized exchange

• Bearman, 1997
• Exchange resources via mediators, depending on
  partner network states
• Tendency to respond cyclically to resource
  promising partner
• cyclicity gt 0
• Not supported in both networks. Probably because
  no need for information exchange via mediators

17
Contagion Exposure

• Burt 1980 Contractor et al., 1990
• Leading to social influence limitation in
  attitudes, knowledge behavior
• Respond to same as other equivalent actors
• Out star gt 0 in star gt 0 mixed star gt0
  transitivity gt0
• Not supported in both networks. Probably because
  contagion process could not develop in the short
  lifetime of networks

18
Cognitive Consistency

• Heider 1958 Festinger, 1957 Cartwright et al.,
  1956
• Drive for balance in cognitions
• Respond via several paths
• transitivity gt 0
• Supported in team networks. Not supported in the
  forum network. In both networks this is due to
  their designs

19
Uncertainty reduction

• Berger 1987
• Reduce uncertainty by gaining
• Attract responses from several others
• In star gt 0
• Not supported in both networks. In the forum
  network the tutor clarified all uncertainties

20
Exogenous factors

• Residual personal tendencies o respond or trigger
  only to actors with pre assigned roles
• For students (1)resp0 (2)trigg0 For tutor
  (3)respgt0 (4)trigggt0
• 1,2 supported for both networks 3 un supported
  for team network, supported for forum network 4
  un supported for both networks

21
(No Transcript)
22
(No Transcript)
23
Thank You

• Questions? Comments? Remarks ?