Research On The Influence Of Group Preference On Performance In Dynamic Group Decision-making

Dynamic group decision-making is more scientific and democratic than static groupdecision-making, because the former takes the communication among decision-makers intoconsideration. However, it s difficult for all the members to reach a consensus and the processtakes much time during the dynamic group decision-making, going against the decisionperformance seriously. According to some empirical research in psychological field, grouppreference affects the group decision performance. In light of that, this paper will discuss therelation between group preferences and decision performance.Firstly, we review the progress and research method of the theory of groupdecision-making. Because of the low efficiency and poor results in dynamic groupdecision-making, the research problem from the perspective of group preferences is posed.Then we make a comment on the concept, characteristic，progress, its performance andinfluencing factors of dynamic group decision-making. Besides the concept and progress ofpreference and7main preference structures in group decision-making are clearly illustrated.Next we elaborate3important preference forms including discrete, continuous and vectorpreferences from the point of opinion dynamics.On the base of discrete preferences, we investigate the relation between the results anddecision time and the ratio, number and distribution of group preferences, combing with thegroup relationships based on scale-free networks. Simulation experiments show:(1) Theevolution results of group preferences depends on the leading preference in the group. Namely,the higher the scale of one preference is, the more likely the group preference will evolve intothis preference.(2)When the number of group preference is larger, the convergence time islonger. But it is proved that they are not proportional.(3)The agents preferences, whose degreeare relatively large in the network of relationship, will domain in the group in steady state.On the base of continuous preferences，we put forward the preference spectrum. Weinspect how preference spectrum affects decision time and dynamic evolution in groupdecision-making. A cellular automaton model is established, in which every decision-makerchoose interacting objects from von Neumann neighbourhood according his confidence leveland change his preference into the weighted average of all his interacting objects preferences.Simulation experiments show:(1)The range of preference spectrum has an big impact on theevolution results. The larger the range of preference spectrum is, the larger the number ofpreference clusters and the longer the convergence time. When the group preferences achieve stability, the ratio of the length of preference spectrum into confidence level determines thedecision time and the number of preference clusters.(3)Whatever distribution of the initialgroup preferences has, the average group preference equals the expectation of initial grouppreferences.