Understanding community: The implications of information flow and social interactions in online discussion groups.

摘要

An increasingly difficult challenge that users of online communities face is how to assess the quality and trustworthiness of content generated by strangers. This is especially true in help and support question-answer newsgroups where little information is displayed in the interface to show an authors' status or reputation within a group or their past behavior in online discussion environments in general. Moreover, different users of online discussion groups behave and interact with information in differing ways.;This dissertation investigates the behaviors that are manifested through various social types when they interact with others in online discussion groups. Mixed methods were employed to ensure the rigor of the research, including (1) segmentation of newsgroup posting behavior by aggregate newsgroups and authors; (2) focus groups with key informants to explore how social norms and the collective worldviews of newsgroup participants shape the ways different social types behave; and (3) a survey addressed to a general population of newsgroup users to gain an understanding of how various social types view their own information behavior and that of others.;This dissertation uses the theory of normative behavior (Chatman 2000) to address the concepts of worldviews (shared understanding of the small world), social norms (what is perceived to be right or wrong in the group), social types (generalized classification of a person or persons) and information behaviors (how a person acts upon given information) exhibited by members of Usenet newsgroups. Implications are suggested for academics seeking a framework for studying online discussion groups, system designers needing advice on how to set up a user community, end users looking for clues on how to decipher and judge information that is being shared, content creators wanting ways to synthesize and preserve important content, and community managers desiring to know how various social types interact with information in order to maximize the experience for the users and the understanding for an organization.