Research On The Employees Counterproductive Work Behavior Based On Catastrophe Theory And Computational Experiments

China is a populous country and has the full labor force. For this reason, Chinese companies focus more on the task performance rather than on the behavioral performance of employees when they enjoy the outcomes based on such low labor cost. However, China has already reached the "Lewis turning point" with social and economic advancement, and the demographic dividend is diminishing gradually. Under this big background, the employees will become Chinese companies’most valuable treasure, and then the impact of their work behaviors especially the Counterproductive Work Behavior (CWB) on organizational performance has become the research the focal point question. The existing researches of CWB are mainly concentrated on the individual level with empirical methods, which mainly include concept connotation, affecting factors and mechanisms. But according to the study is further developed, the increasing nonlinear phenomena of CWB have caused wide public concern. And the control object of CWB is not just individual employee, but is the employees’community which will interact with each other under the complex environment. These are the problems that need to be solved urgently at present.This paper conducted the studies on the CWB of individual and the group from linear to nonlinear, statistic to dynamic and individual to group based on Catastrophe theory and Computational Experiment. At first combining the mutational characteristic of CWB with the previous empirical models, the individual CWB Cusp Catastrophe model is qualitatively built by Element Catastrophe Theory, which under two kinds of dependent variables:personal emotion and situational pressure. And then the Stochastic Catastrophe Theory and Cuspfit are introduced into the research, the thesis demonstrated the Cusp Catastrophe Model by nonlinear empirical study through the data collected from a telecom manufacturer in Wuhan.Using the hypothesis testing of the CWB catastrophic mechanism as a foundation, this chapter mainly focused on the voluntary turnover which is a special form of CWB. Then a parameter solving method of catastrophe model has been designed by qualitative simulation methods and fuzzy math. The range of parameters has been solved through the case of employee turnover in a representative metal processing enterprise in Hangzhou, and finally the results were used to quantitatively measure the voluntary turnover of employees in the economic transition period by a series of numerical experiments.The employees are of high complexity, so the affecting factors of employees’ behaviors include not only two major control variables. There also have some non-critical variables in management practice. A new concept of disturbed mutation was defined to describe the random disturbances. And then the QSIM algorithm was added to individual catastrophe model of CWB, so the strict catastrophe model can be translated to the dynamic and fuzzy Semi-qualitative cusp model. Finally, the Semi-qualitative cusp model is used to analyze the dynamic process of interaction between the CWB and the affecting factors.An employee is not an isolated entity in the organization, but exists in the groups. The key to CWB management is the control of the evolution of group behaviors. Computational Experiments Theory provides new instrument to solve the self-organization and dynamic evolution of collective CWB. At first, the two distinct cognitive systems which under the influence of rational and irrational factors were built based on the qualitative analysis of employees’ psychology. Next an asymmetrical game model was constructed by the Prospect Theory and Evolutional game theory. Finally, the Multi-agent simulation technology is applied to the program of the simulation module of CWB. Then through specially-designed experiments and a typical case of Chinese OEM, we study the control effect on formal and informal constraints which are used to administrate the CWB. These virtual experiments analyze the influence on CWB which are derived from personality traits, management mode, reward and punishment system, organization ethics and remedial measures.A new way of using empirical data to drive computational experiments from the perspective of Bayesian network has been studied. It expands the mechanism of the individual CWB out into the evolutional research of collective CWB. Based on the classical empirical model and real data of Chinese manufacturing enterprises, the decision tree of CWB can be acquired by structure learning and parameter learning of Bayesian Networks. And then the Bayesian Network Data driven Computational Experiment Model is constructed through the combination with social network theory and social exchange theory. Finally, a series of virtual experiments analyze the influence on collective CWB which are derived from group size, the scope of communication, node degrees, antecedent variables and organizational structure. Simulation studies also provide some implications in the real environment for decision making during the control measures of collective CWB.