信任不对称可以合理存在吗?——基于Agent的仿真实验研究

Does Trust Asymmetry Exist Reasonably?A Research based on Agent-based Modeling and Simulation

传统信任研究往往默认一个隐含假设——信任在组织中对称存在,而忽视了信任不对称的普遍存在,或者简单将信任不对称等同于不信任加以排斥。本文借助基于Agent的仿真实验研究,对信任不对称本身是否普遍存在,以及信任不对称在信任均值、同事监督与团队绩效的关系中所发挥的调节效应进行了仿真验证研究。结果证实,信任不对称确实无法根除,但并不必然带来团队绩效的负面效应;信任不对称相对于"过度信任"和"信任缺失"可以合理存在,且能够带来较优的团队绩效。

Whether trust asymmetry can exist reasonably, the topic in this research heads from the following reasons. First, trust asymmetry is one of the unavoidable phenomenons but extensively exists deserving no attention in a long time. Second, the contradictory conclusion reflected by the relationship between trust and performance has caused the necessity to deeply explore whether trust asymmetry can exist reasonably, which is also included in the most important topics facing trust asymmetry. Our research believes that trust asymmetry challenges the traditional trust theory and researchers must envisage it. It is also important to know that asymmetry trust exists in suitable sit- uations. Our research used Agent-based modeling based on the core theory CPOP and simulation as the major methodology, and the formula to calculate trust asymmetry as follows. ∑j=1^k[（∑i=1^r[｜Tab-Tba｜]/r）i]/k In this formula, Tab represents A＇ s trust in B While Tha shows B＇ s trust in A. r represents the number of pair group in the project team, and k is the number of this team. We first calculate the degree of trust difference between each A and B, then average these difference scores for each person and later across the whole team mem- ber. This formula is adapted from the research of Van der Yegt et al. ＇ s （2010） and De long ＆ Dirks（2012）. The way to approach a solution for our main topic is listed as follows. First, we have constructed the computa- tional model of dynamically interaction trust. Then, we set up different initial trust level and observed how trust a- symmetry emerges by running the model. Again, we based on the existed relationship among mean level, peer mo- nitor and trust asymmetry, we further the modeling and observed the range of trust asymmetry which could bring op- timum team performance result. In particular, we added trust rule space in the validated CPOP model, and initial the trust level for each Agent. After running thousands of times, we got the final dataset. The mainly findings shows： we confirm that trust asymmetry definitely exists everywhere, and the simulation methodology helps to restore the real phenomenon. Besides, the lower level of trust asymmetry and higher level of monitor consensus together determine the optimum team performance result. Therefore trust asymmetry is not as simple as distrust. It does not always bring negative performance. It further showed the advance in predicting by simulation, and made sure the importance of peer monitor in maintaining the level oftrust. Peer monitor constitutes the informal control mechanism in organization and transfers the message that＂you have to do like this＂. When ev- ery employee tries to meet the demand of organization and is told that othersalso receive the same orders, the trust level between colleagues will increase. Peer monitor removes those trust obstacles in order to building trust among colleagues and lessens opportunism behavior. And reasonable existence of trust asymmetry means trust average does not always work out. The contribution of our article to the existed research lies on that, first, trust asymmetry as a newly developed filed grew out of trust research, our research steps first to make this filed go further. Second, based on Agent-based modeling and simulation, we constructed a dynamically trust model, which helped us observe the emerging process of trust asymmetry and its impacts. The methodology showed its helpfulness and mayenlightened future research. Third, the research on whether trust asymmetry can exist reasonably showed us the peer monitor played important role in maintaining trust level.