主管:中国科学院
主办:中国优选法统筹法与经济数学研究会
   中国科学院科技战略咨询研究院

中国管理科学 ›› 2016, Vol. 24 ›› Issue (10): 164-170.doi: 10.16381/j.cnki.issn1003-207x.2016.10.019

• 论文 • 上一篇    下一篇

个体信任认知对博弈合作行为的影响

郑君君1, 李诚志1, 韩笑2, 董金辉1   

  1. 1. 武汉大学经济与管理学院, 湖北 武汉 430072;
    2. 武汉理工大学管理学院, 湖北 武汉 430070
  • 收稿日期:2015-10-02 修回日期:2016-04-01 出版日期:2016-10-20 发布日期:2016-12-27
  • 通讯作者: 郑君君(1966-),女(汉族),湖北松滋人,武汉大学经济与管理学院教授,博士生导师,研究方向:博弈论、个体行为研究,E-mail:99zhengjunjun@163.com. E-mail:99zhengjunjun@163.com
  • 基金资助:

    国家自然科学基金资助项目(71371147);湖北省教育厅哲学社会科学研究重大项目(16zd002)

The Effects of Individual Trust Cognition on Cooperative Game

ZHENG Jun-jun1, LI Cheng-zhi1, HAN Xiao2, DONG Jin-hui1   

  1. 1. School of Economics and Management, Wuhan University, Wuhan 430072, China;
    2. Management School, Wuhan University of Technology, Wuhan 430070, China
  • Received:2015-10-02 Revised:2016-04-01 Online:2016-10-20 Published:2016-12-27

摘要: 将信任认知视为局中人关于对方选择合作的信念,探讨在一次性囚徒困境、有限次重复囚徒困境以及无限次重复囚徒困境中,信任认知对局中人策略选择与系统均衡的影响。为了体现信任认知的动态性,以及降低信任预测的主观性,将动态信任预测认知模型引入到重复博弈的分析框架中。研究表明:当个体的总体信任度超过某一临界值时,合作现象将会在系统中涌现并维持。此外,局中人的本轮满意度、历史直接信任度、直接信任度,间接信任度以及活跃度等因素均对其合作行为产生一定的影响。

关键词: 认知模型, 总体信任度, 直接信任度, 间接信任度合作

Abstract: Most of the current researches on trust and cooperation have been studied from the perspective of behavior or from the perspective of outcome. The former takes the trust equal to the individual cooperative behavior, and the later abstracts the trust as an impact factor of player's expected outcome. However, with the development of cognitive psychology, more and more scholars have found that trust belongs to psychological cognitive process essentially, and trust can lead to cooperation, but cooperation cannot explain the degree of trust completely. Hence, no matter studying the trust and cooperation from the perspective of behavior or from the perspective of outcome cannot match our current comprehension of the trust. Against this background, trust is portrayed from the cognitive perspective by the cognitive psychology and the trust is regarded as a faith that opponent will choose cooperation in the interaction.
Dilemma game is taken as example and faith learning and updating is used to portray trust. By using dynamic trust forecasting cognitive model, equilibrium in one-time prisoner's dilemma game, finitely repeated prisoner's dilemma game and infinitely repeated prisoner's dilemma game have been studied, and thus the evolution of cooperation in short-term relationship, the medium-term relationship and long-term relationship have been observed. According to the results, in short-term relationship and medium-term relationship, when the trust degree exceeds a critical value, cooperating with opponents becomes the dominant strategy of the players, and cooperation turns to be the equilibrium of the game system. When players are in long-term relationship, whether the current round satisfaction degree, historical direct trust degree, direct trust degree, or indirect trust degree, and the active degree of the players and other factors have a certain impact on their cooperative behavior at equilibrium. At last, a numerical example has proved this study.
According to our study, the generation and maintenance of cooperation can be promoted by improving trust degree between individuals. Generally speaking, approaches to enhance the trust degree between individuals are as follows:increasing indirect trust degree by establishing and maintaining a reputation, improving satisfaction degree by improving personal performance in each interaction. For those individuals who already have relatively good reputation, it is wise to enhance their active degree of the social system to endow more weight to their direct trust. In all, This study can help us understand the impact of trust on cooperation from the perspective of cognitive psychology.

Key words: cognitive modeling, over all trust degree, direct trust degree, indirect trust degree

中图分类号: