Schneider and Reichers (1983) delineated three perspectives on the formation of climate: structuralist, attraction-selection-attrition (ASA) or homogeneity, and social interaction. Although it is sometimes considered part of the structuralist perspective, we also discuss separately the importance of leadership as well as the impact of an individuals' immediate workgroup.
Structure and Practices
In the structuralist perspective, climate purportedly arises out of structural characteristics of an organization. With its roots in Lewin's (1951) field theory, this approach assumes that organizational characteristics such as size and structure establish a common reality that provides the basis for shared perception. The person is analytically separate from the social context (Roberts et al., 1978). Support for this perspective has rested on research showing that climate perceptions are related to structural variables such as size, centralization, structure, and hierarchical level (e.g., Payne & Pugh, 1976).
More consistent with current definitions of climate, the set of policies, practices, and procedures of the organization are
the features that provide the basis for shared perceptions to emerge. However, merely introducing and implementing a set of practices around some strategic focus is not sufficient Unless the practices are designed and implemented in such a way as to create a strong situation (Mischel, 1973), idiosyncratic psychological climate perceptions are likely to emerge (Ostroff & Bowen, 2000). Organizational practices, such as human resource management (HRM) practices, are often viewed as communications from the employer to employee (Rousseau, 1995; Tsui, Pearce, Porter, & Tripoli, 1997). Yet two employees can interpret the same practice differently because there is considerable variance among employees in the sense-making strategies they use in interpreting these messages (Guzzo & Noonan, 1994). Differences in category systems or cognitive maps can exist across people (Kelly, 1955). To the extent that the situational stimulus is ambiguous or unclear, multiple categorization is likely (Feldman, 1981) and different people are likely to use different cognitive categories to attend to different aspects of the situation, making subsequent attributions different. This can result in divergent views about appropriate behaviors (Rousseau & Wade-Benzoni, 1994).
On the other hand, collective sense-making can occur when practices are designed to induce a strong situation, regardless of the type of practice implemented. For example, when practices represent a coherent and internally consistent whole; are made very visible; are communicated widely and clearly; are administered consistently throughout the organization; and are fair, legitimate, and valid, a collective perception of climate based on these practices is more likely to emerge (Ostroff & Bowen, 2000). These metacharacteristics of the practices help reduce ambiguity and enhance clarity of interpretation in the setting, thereby allowing for similar cognitive maps to develop across people so that the context and appropriate ways of behaving are understood. A strong process of delivering practices creates the elemental content and this content is shared because interpretations are consistent across people. One implication is that in today's world of virtual organizations or virtual teams, in which interaction and communication are often limited, climate could be perceived similarly if the practices are unambiguous and delivered in a strong manner (although climate in general may be less useful as an explanatory variable with such distributed work arrangements).
Homogeneity
This approach is based on the ASA process (Schneider & Reichers, 1983) in which individuals are attracted to and want to join organizations that have similar attributes to their own
views and attributes. Selection procedures attempt to ensure that the applicants hired fit the organizational context, and people tend to leave organizations when the work context does not fit their personal characteristics. As a result, an organization is likely to consist of very similar people (Schneider, 1987). These effects may be furthered by the socialization processes that can change new organizational members' personal attributes, goals, and values in the direction of those of the organization (Fisher, 1986; Ostroff & Rothausen, 1997). Due to this homogeneity process, individuals should perceive the organization similarly and sense-making should not yield idiosyncratic interpretations, but should yield agreement (Schneider, 1983).
Little empirical work has explicitly tested the relationship between homogeneity and the development of shared climate perceptions. Early work, largely in educational settings, showed that personality was related to organizational climate ) perceptions (e.g., J. R. George & Bishop, 1971; Stern, 1970). Recently, group cohesiveness has been positively related to agreement on climate perceptions (Naumann & Bennett, 2000) and needs have been related to collective climate membership (Young & Parker, 1999).
Social Interaction and Communication
The third approach to the emergence and formation of organizational climate is based on social interaction, with roots in symbolic interactionism. Shared perception and meaning evolve from communications and interaction patterns among members of the same group. Work contexts are created by the individuals within them—that is, "the people make the place" (Schneider, 1987); and at the same time, the context and individuals' interpretations of it have a large influence and impact on behavior and responses (e.g., Ashforth, 1985; Morgeson & Hofmann, 1999; Schneider & Reichers, 1983).
Overlapping schemas or cause maps across people can be facilitated through social exchange and transactions among employees. In such a way they can agree on the appropriate aspects of the environment to attend to, and on how to interpret these aspects and respond to them appropriately (Weick, 1995; Wicker, 1992). As explained by Morgeson and Hofmann (1999), within any collective, individuals are likely to meet one another and interact. Each interaction results in a discrete event, and subsequent interactions are termed event cycles. The structure of any collective group can be viewed as a series of ongoing events, activities, and events cycles among the individuals. These interdependencies and interactions among individuals over time can result in jointly produced responses, and it is this structure that forms the basis for the eventual emergence of collective constructs. Group
Emergence of Shared Meaning and Perceptions 581
members construct the meaning of organizational events from repeated social interactions and it is these interactions that are likely to result in conformity (Ashforth, 1985).
Some recent empirical studies have shown that collective climates (defined through clusters of people with similar perceptions) are related to formal workgroup membership (e.g., Jackofsky & Slocum, 1988; Young & Parker, 1999). However, other research has found little or no relationship between collective climates and workgroup membership, regional membership, job type, or hierarchical level (e.g., Patterson etal., 1996; Gonzalez-Roma etal., 1999). One explanation for these contradictory findings is based on informal interaction groups, which may have more influence on the development of shared climate perceptions than structurally imposed interaction groups. Indeed, research has shown that informal interaction groups attached similar meanings to the organizational context (Rentsch, 1990) and that individuals' climate perceptions were more similar to those of others with whom stronger communication ties were maintained (Fink & Chen, 1995). Clearly, more research is needed in this area. For example, researchers should investigate the extent to which interactions and interdependencies actually occur in structurally imposed groups, and whether it is these interactions that form shared perceptions. Likewise, given that exposure to and participation in various work practices can result in more positive perceptions about the practices and the organization (Katz & Kahn, 1978), much more research is needed in determining how individuals shape the climate and affect climate perceptions.
Leadership
Finally, shared perceptions may result through leadership processes. Original conceptualizations of climate focused largely on the role of the leader in creating climates (e.g., Lewin et al., 1939; McGregor, 1960) and experimental studies showed that climates became increasingly differentiated over time in a manner consistent with a leader's style (Litwin & Stringer, 1968). Leaders or supervisors serve as interpretive filters of relevant organizational processes, practices, and features for all group members, contributing to the development of common climate perceptions (Kozlowski & Doherty, 1989). By exposing employees to the same policies, practices, and procedures, leaders act as "climate engineers" (Naumann & Bennett, 2000) or "meaning managers" (Smircich & Morgan, 1982).
For example, in a study of service climates, the importance of service to managers was related to employee perceptions of the service climate (Borucki & Burke, 1999). This is likely due to the fact that managers and leaders are largely responsible for communicating meaning (Schein, 1992). Leaders explicitly
582 Organizational Culture and Climate
and directly communicate their own interpretations and, in conjunction with interacting with most members, will be able to introduce a common interpretation among unit members (Rentsch, 1990). Indeed, in a recent study, agreement among workgroup members on climate was stronger when the supervisor was seen as visible in implementing organizational procedures and enforcing policies (Naumann & Bennett, 2000). Moreover, several studies have indicated that high-quality-exchange relationships with a leader are related to climate perceptions and that subordinates in high-quality-exchange leader-member relationships had greater within-unit consensus on climate perceptions (e.g., Kozlowski & Doherty, 1989; Scott & Bruce, 1994). The causal direction of the relationship between climate and leader-member relationships, however, is not yet known (Cogliser & Schriesheim, 2000). f Furthermore, the literature on social influence and authority indicates that individuals are willing to adjust their behavior in response to the inductions of some influencing agent (Kelman & Hamilton, 1989). The influencing agent, or in this case leader or supervisor, can exert influence and induce uniform behavior through informing the person about the nature of the situation and what is at stake for him or her in this situation, explaining practices, policies, and procedures and the consequences for adhering or failing to adhere to them, and enacting the organizational practices. Not only can a leader create shared perceptions by serving as a filter of and model for organizational practices, policies, and procedures, but through influence can induce uniform responses in accordance with these practices, which are then interpreted by members in forming perceptions of climate. The role of the leader in forming climate perceptions is an area ripe for research.
Workgroup Influences
As noted earlier, the aggregate level of analysis refers to any higher level (e.g., division, function, unit). The most immediate and proximal level is likely to have the greatest influence (Katz & Kahn, 1978; Rousseau, 1985); hence, processes within an individual's immediate workgroup or team should help in the formation of shared cognitions.
Recent work in the area of teams has highlighted the concept of shared mental models as an underlying mechanism for team effectiveness, enabling team members to respond appropriately and effectively in their work environments (Kozlowski, Gully, Salas, & Cannon-Bowers, 1996; Marks, Zaccaro, & Mathieu, 2000). For example, using a technique called concept mapping, Marks et al. showed that leader communication in the form of transmitting, exchanging, reporting, and passing on information about the task and the work
environment as well as training focused on team interaction were related to the development of shared mental models about how the work system and environment operate. Similarly, Hofmann and Stetzer (1996) showed that the group process (e.g., sharing information, coordinating efforts, interdependence) was related to shared perceptions of a climate for safety, and that both group process and climate were related to individual-level and group-level safety outcomes.
While we acknowledge that the factors discussed previously (structure, ASA, communication, leader roles) are likely to have effects on the emergence of shared perceptions within a work unit similar to the effects they have for any level of analysis, the proximal nature of the immediate workgroup makes it worthy of separate mention. There are also other workgroup processes, such as task characteristics and workgroup structure, and common workgroup experiences such as successes and failures (Marks et al., 2000), that are likely to influence the development of shared perceptions among workgroup or team members. In this regard, researchers might find the perspective offered by structuration theory (Poole, 1999) useful. Structuration is a general theory that can be applied to any emergent or developmental phenomenon in groups and that acknowledges the role of action, group interaction, and their products. Essentially, this theory focuses on understanding the structuring process, or the explicit and implicit rules and resources that members use to generate and sustain the group system and that serve as guides for action.