Midlevel Administrators and Their Intent to Leave
The "morale" of faculty and staff is often used to characterize the quality of academic life within a particular campus or institution. Typically those commenting on morale have an intuitive sense that an individual's morale is "high," or the morale of the administrative staff is "low," or that the faculty's morale has "plummeted." These comments often refer broadly to the level of well-being that an individual or group is experiencing in reference to their worklife (Johnsrud, 1996). We frequently hear references to "morale," but it is not a well-defined or precisely measured concept.
Nor is it clear what effect morale has on behavior. Again, it makes intuitive sense that the higher the morale, the higher the performance, but there is little empirical data. Lindgren (1982) has argued that administrators increase their effectiveness when they are personally affirmed. Similarly, Johnsrud and Rosser (1997, 1999a) have shown that morale is related to administrators' intent to leave their positions. Despite the seeming efficacy of morale, however, there is little agreement as to its definition or understanding of its impact. The purposes of this study are:(1) to define the construct of "morale" empirically and examine it within a broader theory of how organizations affect individuals and (2) to further investigate its construct validity by proposing and testing a multilevel model concerning the impact of morale on administrators' intentions to leave their positions.
Conceptual Framework of the Study
Research on organizations has led to the formulation of theories about the structures and processes within an organization that can make a difference in members' views, beliefs, attitudes, and behaviors (e.g., morale, effectiveness, turnover). Proposed theories often become problematic, however, when researchers attempt to model the actual detail and richness of real organizations because of the complexity of interrelationships that comprise organizational processes. Researchers often lack the ability to isolate, define, and measure some of the important variables that are needed to fully test the theory (Hofstede, Nueijer, Ohayv, & Sanders,1990; Hox, 1995; Marcoulides & Heck, 1993). Though we acknowledge that organizations are socially constructed by their members, we believe that at least some aspects of organizations can be measured and yield important information about influences on individual behavior within organizations.
Theoretical development has also been hampered by a lack of analytical tools that can adequately address the hierarchical structures of organizations; that is, individuals are nested within subunits that are nested within larger units. Previous research conducted at a single level of analysis offered few options for modeling this complexity. Ideally, multilevel theories of organizational processes should specify which variables belong to which level, and which direct effects and cross-level interaction effects can be expected (Hox, 1995). Cross-level interaction effects between the individual and the organizational context require the specification of some process within individuals that causes those individuals to be differentially influenced by certain aspects of the context (Hox, 1995; Hummel, 1972; Stinchcombe, 1967). The common denominator in these theories is that they all postulate one or more social psychological variables that mediate between individual variables and contextual variables (Bluedorn, 1 982; Hox, 1995). Our concern in this study is to further our understanding of how organizational contexts affect individuals, their perceptions, and their subsequent behavior--in this case, turnover behavior.
Organizational theorists (e.g., Bluedorn, 1982; Price, 1977) have developed a model of the process producing voluntary turnover composed of structural, economic, and social psychological variables. …