关键词: 在线知识社区, 协作冲突, Wikipedia, 层次回归分析

Abstract: The online knowledge community (OKC) advocates a user-generated content (UGC) model of “anyone can edit”, providing opportunities and platforms for the civilian population to participate in knowledge production, successfully subverting the traditional professional-generated content (PGC) model and opening a new era of knowledge production. However, with the development of OKC, the phenomenon of collaborative conflict is becoming more and more common. Individuals need to spend a lot of time and energy to resolve collaborative conflicts, which reduces the willingness of individual participation and the quality of knowledge construction. At the same time, a large number of collaborative conflicts lead to increased coordination and management costs, which seriously hinders the development of the community. The main problem currently confronted with OKC is how to effectively prevent, solve and control collaborative conflict, and the key to solving this problem is to find out the reasons behind collaborative conflict. Therefore, it is of great practical significance to study the influencing factors of collaborative conflict in OKC.The pluralism of conflict determines the complexity and diversity of the influencing factors of conflict. Based on combing and summarizing the existing literature, most scholars follow the ABC (antecedent-behavior-consequence) research ideas, and take conflict as the mediator variable or moderator variable between the antecedent variable and the outcome variable. Few scholars specifically study the influencing factors of conflict. In addition, most of the existing researches are directed towards traditional teams, while the researches on virtual teams are scarce, and the research models are relatively simple and lack a systematic theoretical framework. Therefore, this paper takes collaborative conflict as the research object, explores the influencing factors of collaborative conflict, and reveals the mechanism of collaborative conflict.The collaborative knowledge construction of OKC is essentially a dynamic process in which a large number of heterogeneous individuals promote cognitive aggregation by cognitive interaction. Based on social cognitive theory, this paper establishes a theoretical model about the collaborative conflict in OKC, which consists of knowledge heterogeneity, group differentiation, anonymity, task complexity and coordination mechanism. Then, we propose the research hypotheses according to the relevant theory and research, which are as follows: There is a positive relationship between collaborative conflict and knowledge heterogeneity, group differentiation. There is an inverted U-shaped relationship between anonymity and collaborative conflicts. Task complexity positively moderates the relationships between collaborative conflict and knowledge heterogeneity, group differentiation, anonymity. Coordination mechanism negatively moderates the relationships between collaborative conflict and knowledge heterogeneity, group differentiation, anonymity.Wikipedia is a multilingual, web-based, free encyclopedia based on a model of openly editable and viewable content, a wiki. Wikipedia, the world's largest online knowledge community, attracts participants from all over the world, and conflicts are common during collaborative editing. Currently, English Wikipedia is the largest of the more than 303 Wikipedia encyclopedias. Therefore, this study selects the English Wikipedia as the research object. In view of the huge number of Wikipedia entries, we are unable to perform full sample processing. Therefore, we select samples by random sampling according to the quality level of the Wikipedia entry. At the same time, in order to ensure that the samples are typical and representative, we only select Wikipedia entries with a quality level above B. We select 400 Wikipedia entries, 100 entries for each of featured entry, good entry, A-class entry and B-class entry. In addition, we screen the samples according to certain criteria, and finally collect 364 valid English Wikipedia entries as samples. We obtain empirical data of each variable (collaborative conflict, knowledge heterogeneity, group differentiation, anonymity, task complexity, coordination mechanism and control variables) through web crawling and manual acquisition, which is used in empirical research to validate research model and hypotheses.This study mainly adopts hierarchical regression analysis to empirically analyses the relationships between collaborative conflict and knowledge heterogeneity, group differentiation, anonymity, and specifically explores the moderating effects of task complexity and coordination mechanism. To avoid the problem of multicollinearity, we follow three steps for data processing. Firstly, we centralize the explanatory variables (knowledge heterogeneity, group differentiation and anonymity) and the moderating variables (task complexity and coordination mechanism). Secondly, we test the main effect of the three explanatory variables on collaborative conflict. Finally, we introduce interaction terms of explanatory variables and adjustment variables into the model, the purpose of which is to test the moderating effect of task complexity and coordination mechanism on collaborative conflict. The empirical study results show that knowledge heterogeneity and group differentiation both have a significantly positive relationship with collaborative conflict, and anonymity has an inverted U-shaped relationship with collaborative conflict. In addition, task complexity positively moderates the relationship between collaborative conflict and knowledge heterogeneity, anonymity. Coordination mechanism positively moderates the relationship between knowledge heterogeneity and collaborative conflict, and negatively moderates the relationship between group differentiation and collaborative conflict.This study also has important practical implications for conflict management in OKC. Firstly, OKC should strive to create an open and inclusive collaborative environment and establish a cooperative relationship based on mutual respect. When there are multiple different views on the same subject, each view is allowed to be expressed equally, avoiding arguments and conflicts. Secondly, OKC should highlight work priorities and strengthens conflict management. Anonymous users are more likely to trigger collaborative conflicts, so OKC should pay more attention to such users. At the same time, when there is a serious conflict in the community, the community should adopt temporary protective measures to force the parties to remain calm and avoid further escalation and deterioration of the conflict. Finally, OKC should strengthen group communication and improve coordination mechanism. On the one hand, OKC should enhance interactive functions (such as developing instant chat tools), solve communication barriers among members, and promote consensus. On the other hand, OKC should improve various guidelines to effectively regulate and constrain individual editorial behavior, improve group collaboration efficiency, and promote knowledge construction.In summary, this paper systematically and deeply studies the influencing factors of collaborative conflict in OKC. This paper not only enriches the theory of collaborative conflict in OKC and makes up for the shortcomings of existing research, but also has important practical significance for further optimizing community platform construction and improving conflict coordination and management mechanism.

Key words: online knowledge community, collaborative conflict, Wikipedia, hierarchical regression analysis