In more recent years, the study of trust has appeared in the field of different disciplines and become a popular research topic. In particular, the issues related to organizational trust have also been an important concept of organizational research over the past two decades. In the past research, trust was considered to be an important factor affecting job performance, satisfaction, organizational commitment, cooperation, and Organizational Citizenship Behavior (e.g., Colquitt et al., 2011; de Pablo González et al., 2014; Frazier et al., 2015; Lau and Cobb, 2010; Saunders and Thornhill, 2003). Then, including leadership effectiveness, organizational justice, and relationships conflict is viewed as the influence factors of trust(e.g., Podsakoff et al., 1996; Costa, 2003; Lau and Cobb, 2010; Colquitt abd Rodell, 2011 ); In the meanwhile, trust is also an important mechanism or conditional factor in human resource management practices and organizational performance, employee attitudes(Tzafrir, 2005; Innocenti et al., 2011), leadership and assistive behavior (e.g., Zhu and Akhtar, 2014), leadership and organizational performance, and organizational citizenship behavior (e.g., Miao, 2014) .According to the social exchange theory of Blau (1964), social exchange has the expectation of the future that there is no immediate return. Rewarding actions would have become an obligation. Only when the obligation is fulfilled and the expectation is satisfied, the interpersonal trust can be established. Therefore, interpersonal trust manifests the dyadic relation of social interaction which establishes the cohesion and mutual trust between members and managers in the organization (Hotho et al., 2012) , including colleagues, supervisors and subordinates, and higher manager and employees. In the interactive network, trustworthy people can develop exchange relations and promote the accumulation of social capital. In the organization, the manager is the creator of trust, and trustworthy managers will influence the exchange relations of subordinates (Whitener et al., 1998). Further, we believe that employees have different perceptions of trust between supervisors and higher level manager. Previous studies have largely examined trust in supervisors and trust in peers, but was less regarding trust in subordinates. However, trust in subordinates as the character or behavior of leader has been considered an important element of high quality of the leader-subordinate relationship (Boyatzis et al., 2006; Cunliffe and Eriksen, 2011).Based on the Role-Set Theory, this study suggest that people are given different role expectations at the position of authority level. For those, key figures, who can influence personal attitudes and behavior should be considered the frequency of interaction and distance with subordinates (Merton, 1957; Ruiz et al., 2011). Hence, the impact of top management on the organization is comprehensive. Different management levels in the organization will cascades dowm to the lower-level employees through the supervisor. (Mayer et al., 2009). According to the social learning theory (social learning theory; SLT; Bandura, 1977) and social exchange theory (social exchange theory; SET), we suggest that supervisor will imitate the leadership style of higher level manager. If team members feeling trusted by supervisors, they tend to have cooperative behavior and attitude . This study agrees with Mayer et al. (2009), that the interaction between the managers and employees of different level are deemed distinction, and employees are affected differently because of hierarchical relationships. Therefore, the impact of higher level manager on the attitude and behavior of employees will mediate by supervisor.In addition, this study suggests that the social dilemma of trust and group size are also a notable issue. Some researches have been tested the effects of trust on employee behavior might be moderated by job demands (Colquitt et al., 2011) or industrial relations climate (Kougiannou et al., 2015). We propose team size as a moderator of the relationship between feeling trusted by supervisors and team cooperation. Larger team size will not only reduce team cohesion, it may also affect the team trust and cooperation (Alnuaimi et al., 2010). In the following sections, we discuss the conceptualization of theoretical constructs and theoretical background, and provide the theoretical basis for the research hypotheses.Theoretical background and hypothesesWork group level of analysisThis study takes the group level as the analysis unit, so we conceptualize subordinates’ felt trust, team cooperation at the work group level. These group-level constructs are treated as “shared unit properties” (Kozlowski and Klein, 2000) through social interaction and exchange between employees of the workforce. Most of studies suggest that cooperation is an individual behavior. In this study, We expect to evaluate the perception of team cooperation at the group level. Eby and Dobbins (1997) has examined the perception of team cooperation in work groups and assessed the relationship between team members’ quality and reciprocity. Next, we presented some theories that help explain why these constructs might emerge from the group-level. First, social information processing theory suggests that the social context of an individual is the source of important clues to construct and interpretate phenomena , and this context guides the predictability of individual’s behavior and the logic of action (Salancik and Pfeffer, 1978). In the same work context, task interdependence leads to common behavioral norms, as well as homogeneous attitudes and preferences (Eby and Dobbins, 1997). Thus, the majority of employees have a shared perception of the behavior of the group and the tendency of event judgement.Second, according to Bandura’s (1977) SLT, he suggests that learning is the process of interaction between the individual and the social environment, and that individual behavior is influenced by observation. Individuals will strive to emulate the behaviors of role models. Therefore, in the working environment, the employee will observe the supervisor or colleague who is worth to be learned to ensure that their behavior is in line with acceptable norms. If there is a commendable figures in the organization, the employee will tend to model their behaviors as much as possible, and the work group members will imitate each other so that they behave similarly (Mawritz et al., 2012; Mayer et al., 2009).Third, the attraction-selection-attrition (ASA) model suggests that individuals are attracted to and selected into an organization based on their preferences that is consistent with their own personal characteristics, and that the organization will select the suitable person during the selection process. Individuals will remain a member of group or organization when their behavioral trends fit with other members, and those that do not fit will leave(Schneider et al., 1995). As a result, members in the workgroup have homogenous behavioral norms and ideas with each other.Although the basis of theoretical frameworks are different and are not particularly concerned with the subject of this research, these can predict the group behavior. We suggest that all three enough to provide a reasonable explanation of the group-level constructs.The concept of trustTrust is established by the interaction between