An exploratory study on message generation in virtual communites and the influence of trust


Student thesis: Doctoral Thesis

View graph of relations


  • Sau Ling Fion LEE

Related Research Unit(s)


Awarding Institution
  • Douglas VOGEL (Supervisor)
Award date15 Feb 2006


The emergence of virtual community has provided a channel for people to communicate and share. Individuals and businesses can use virtual community to achieve purposes such as collecting comments from customers and sharing personal feelings. This can usually be carried out through posting and reading messages in discussion boards, including information exchange messages, emotional support messages, and overhead messages. In view of this, this study investigates which community type individuals or businesses should develop in order to obtain the desired message type. This research also investigates how trust influences the occurrence of different message types in different community types. A further aspect is ascertaining which community type associates with more information exchange messages and which obtains more emotional support messages. This study first proposes four virtual community types based on different combinations of technology level and message intensity of the community. The four virtual community types proposed are: Activity Focused Community, Strong Technology Supported Community, Emperor Community, and Declining Community. This classification scheme is then used to classify 200 virtual communities in the Internet. Results show that this classification scheme can help in matching each selected community to one of the community types proposed. This suggests that the comprehensiveness of the classification scheme is appropriate. Taking into consideration that the Emperor Community type can be regarded as the integration of the other three community types, this study focuses on conducting detailed investigation on the remaining three virtual community types. Messages have been extracted from discussion boards of six virtual communities and these messages have been used to perform a protocol analysis according to a predefined coding scheme. The protocol analysis aims at finding out the associations between different community types and message types, and the influence of trust on these associations. Statistical results largely support that information exchange message is the most popular message type in all three community types. Emotional support message is the second most dominating message type in Activity Focused Community and Strong Technology Supported Community but it is the least frequent message type in Declining Community. Instead, overhead message is the second most dominating message type in Declining Community while it is the least popular in the other two studied community types. Results also supported that Strong Technology Supported Community obtains the most emotional supported messages compared to the other community types; they partially support that Activity Focused Community has the most information exchange messages among all community types. Statistical results relating to influence of trust are quite diversified. Based on the findings of protocol analysis, interviews were conducted on six participants of virtual communities. Interview results are largely consistent with those in protocol analysis. Detailed data regarding influences of trust and distrust were collected, such as whether distrust occurs more frequently than trust in virtual community. However, due to the complexity of trust, there is no conclusive result on the impact of trust on the virtual community.

    Research areas

  • Trust, Computer networks, Social aspects