Shared Meaning and Perceptions

Shared meanings and perceptions are the foundation of organizational-level or unit-level culture and climate. We now discuss a variety of issues associated with the methods used to establish the extent of shared meaning or convergence of perceptions.

Demonstrating Agreement

In the culture area, meaning has most often been assessed through qualitative studies. Culture researchers elicit interpre­tations of what the organizational context means to employees (e.g., Langan-Fox & Tan, 1997) and from these assessments summarize meaning into some aggregated qualitative descrip­tion of the culture. Thus, although studies of culture have as­sessed interpretations of events, the qualitative method does not well allow for objective comparisons across units or for direct assessment of the extent of agreement.

In contrast to the culture literature, climate researchers have devoted considerable attention to measuring and docu­menting the degree to which organizational members share perceptions of the organizational climate (Schneider, 2000). The most common procedure is to use a mean or aggregated score across individuals within the same unit to represent a higher level climate. However, procedures and criteria for de­termining whether the mean scores can be interpreted as an in-dicatorof the higher level climate have been a source of debate (Patterson, Payne, & West, 1996). One criterion rests on the demonstration of between-group differences between units on their mean scores (cf. Glick, 1988; L. R. James, 1982). A second criterion rests on the demonstration of within-unit agreement or consensus to show that climate, conceptualized and operationalized at the individual level, is functionally

578 Organizational Culture and Climate

isomorphic to another form of the construct (e.g., organiza­tional climate) at the higher level (cf. Danserau, Alluto, & Yammarino, 1984; L. R. James, 1982; L. R. James, Demaree, & Wolfe, 1984). While debate continues about the most ap­propriate indices, researchers have generally agreed that some form of within-unit agreement or consensus in responses and sufficient between-unit variability be demonstrated in order to justify using the mean score to represent a higher level climate (Klein etal., 2000).

A related issue pertains to the referent or focal point for assessing climate. Many assessments of climate have had the focal point of measurement as the individual (e.g., I per­ceive . . .). However, some researchers (e.g., Chan, 1998; Klein, Conn, Smith, & Sorra, 2001) have argued that rather than measure an individual's own climate perceptions, we should assess how an individual believes most people in the organization perceive the climate and whether there is within-unit consensus in such beliefs. Thus, conceptualization of the climate construct is still at the level of individual per­ception, but the referent of the content is changed to the unit level (from self to others). Finally, consensus or agreement can be imposed by having individuals discuss and come to agree­ment on the climate as a group. The resulting score is then used as the indicator of the climate, making statistical indices of agreement unnecessary. More research is needed to determine the implications of this shift in focal point and the use of group-based agreement techniques for the construct meaning of climate across levels of analysis.

(Dis)agreement

The absence of shared perceptions has been addressed in both the culture and climate literatures. For example, the deviance model (Martin, 1992) or the dissensus model (Trice & Beyer, 1993) of culture highlights disagreement or lack of consensus. However, there is debate as to whether deviance or dissensus in an organization indicates that a culture exists, a fragmented culture exists, or no culture exists. In the climate literature, large variability in perceptions among members indicates that aggregated perceptions do not adequately represent a construct of climate at the higher level (e.g., L. R. James, 1982; Klein etal., 2001).

Empirical studies of climate have often found that although agreement on climate may be adequate from a methodological standpoint for the study as a whole, there is still considerable variability in perceptions, and some groups or organizations in the sample have less than adequate agreement on climate perceptions. Thus, additional models for addressing the link between individual perceptions and aggregate constructs have been suggested. For example, in a dispersion model (e.g.,

Chan, 1998; Kozlowski & Klein, 2000), the degree of vari­ability of responses (e.g., an r_wg score) can be an important variable in its own right (not only a justification for an aggre­gate score), independent of the level of the content of climate (e.g., mean climate on some climate dimension). To the extent that greater homogeneity in perceptions of climate is present, collective perceptions and responses should be more uniform and organizational-level relationships can emerge and be meaningfully examined (Ostroff & Bowen, 2000).

Collectivities

Recently, cluster analysis has been used to form subcultures (e.g., Hofstede, 1998) and collective climates, which repre­sent clusters of employees who perceive the organization similarly (e.g., Jackofsky & Slocum, 1984; Patterson et al., 1996; Gonzalez-Roma, Peird, Lloret, & Zornoza, 1999). Be­cause subcultures and collective climates are formed by grouping people based on the similarity of their perceptions, the agreement problem is essentially solved (Payne, 1990). Nevertheless, collective climates may not be representative of a meaningful organizational construct, but rather may be statistical artifacts, particularly if they do not correspond to any defined, formally or informally structured collectiv­ity such as workgroups, divisions, or hierarchical levels (Patterson et al., 1996; Payne, 1990). The foundation of this argument is relevant to the unresolved and underresearched question about how individuals come to share similar percep­tions of the work environment (Schneider & Reichers, 1983; Young & Parker, 1999).

Date: 2016-03-03; view: 643