虛擬社群中誰比較可信任？

Who Is More Trustworthy in Virtual Community?

在匿名的虛擬世界中，誰比較可信任？這問題對網站經營者、參與者與研究者而言，都是重要而難解的。心理學家們認為：過去行為是預測未來行為的最好指標。或許，在匿名的虛擬世界中，這句話仍然有效。本研究目的即嘗試：將MyAV虛擬社群以往互動的關係資料，轉換成社會網絡矩陣後，以個體在社會網絡的中心性位置，來預測個體的可信任度。結果顯示，「外向程度中心性」可顯著地預測可信任度的專業構面（β= 0.28, p < 0.01）;而「內向親近中心性」則可顯著地預測可信任度的正直構面（β= 0.35, p < 0.01）。此方法較諸社群網站所提供的發言次數等活躍度指標，其預測力相對高出許多。

It is important for participants and owners of virtual communities to find out which participants are trustworthy. However, this challenge has not been adequately addressed in practice and /or in research. Following a well-known tenet that past behavior tends to predict future behavior, we applied social network analysis to examine how participants' past interactions in the MyAV virtual community predict trustworthiness. According to the social network indices derived from a social network matrix, which represents past interactions in the community, we found (1) the "out-degree centrality" index significantly predicts the competence dimension of trustworthiness (β= 0.28, p < 0.01); while the "in-closeness centrality" index significantly predicts the integrity dimension of trustworthiness (β= 0.35, p < 0.01). Foremost, both indices outperform other conventional variables (e.g., frequencies of posting messages in the virtual community) in predicting trustworthiness.