Abstract: As the pressures deriving from the need to develop new business models in dynamic, uncertain and complex environments and the need for innovation are increasing, many organizations find the use of groups efficient and productive. One type of behavior that may contribute to the effectiveness of groups is group members’ citizenship behavior. Focusing on groups, citizenship behaviors need to be investigated at the group level. With respect to understanding group organizational citizenship behaviors (GOCB) in organizational settings, Work-group size is a crucial factor to be included. Many studies have demonstrated that individual organizational citizenship behaviors are strongly related to leadership. China has historically been regarded as a case of collectivism. In general, the group orientation at work is emphasized in collectivist cultures. Therefore, GOCB is likely facilitated by Chinese culture. It seems that investigating the influence of group size and leadership on GOCB is a natural step to take. GOCB and leadership are targeted as group phenomena, and therefore should preferably be measured at that level. Based on the referent-shift consensus model, GOCB and leadership scale was employed to the survey. Data were collected from a survey of 501 employees at 37 organizations in mainland China. The hypotheses identified the group as the unit of analysis, so all variables are aggregates of individual responses to the group level of analysis. Justification for aggregation is provided by the demonstration of agreement within settings rather than differences across groups. A well established procedure to assess the factor structure of a scale at multiple levels of analysis is multilevel confirmatory factor analysis which was applied to the data for GOCB and leadership. Results showed that the factor loadings of the items were stronger at the between (i.e., group level) than the within (i.e., individual) levels of analysis. These findings indicated that GOCB and leadership can be analyzed at group level. After fitting the multi-level confirmatory factor analysis model, research hypotheses were tested at the group-level. Results from the sample confirmed the negative relation between group size and GOCB (p < .001) and positive relationship between C and M dimensions of CPM model and GOCB (p < .001). There is no significant link between P dimension of CPM model and GOCB (p > .05). With increases in group size, the psychological distance between individuals can increase. It would be expected less exhibition of GOCB as the psychological distance between group members increases. Chinese culture focuses on moral leadership. C dimension of CPM model implies considerable moral standing on the part of the leader. Without this, it is difficult to conceive how followers could be motivated. Generally speaking, Chinese leaders are perceived more people-oriented than task-orientated. This could be the reason for the M’s positive effects on GOCB and P’s weak influence on GOCB. The outcomes of this study are relevant for management. Faced with the complex and uncertain nature of today’s workplace, managers should become more aware of the effects of their moral practice and consideration on the willingness of employees to engage in GOCB. Administrators should not forget that group size is one of the key elements to encourage employees’ GOCB.

Key words: Group organizational citizenship behavior, Leadership, Multi-level confirmatory factor analysis

摘要： 以37个工作群体中的501位员工为对象，探讨群体规模与领导行为对群体组织公民行为的影响。运用多层验证性因子分析技术对调查数据进行了检验，并在此基础上进行了最优尺度回归和偏最小二乘回归分析。结果表明，群体规模越大，群体组织公民行为的水平越低。领导的个人品德和团队维系行为对群体组织公民行为具有显著的正向影响作用，而领导的工作绩效行为对群体组织公民行为的影响作用不显著。

关键词: 群体组织公民行为, 领导, 多层验证性因子分析