Content area
Full Text
Structural equation modeling was used to evaluate the effects of three forms of attitudinal commitment on employee turnover. We used a traditional dichotomous measure of organizational departure but applied the PRELIS technique to resolve statistical problems. A test of eight competing versions of a turnover model using data collected from employees in an aerospace firm indicated that the forms of commitment affected turnover indirectly through a latent variable measuring withdrawal tendencies. The model including this variable resembles cognitive psychologists' descriptions of human mental processes that emphasize actors' general orientations and a lack of distinction making in everyday life.
The idea that organizational commitment is in some way related to employee turnover has received considerable attention in recent years. Progress in documenting organizational commitment as a correlate of withdrawal process variables is evident from meta-analyses of the research linking these concepts. Cotton and Tuttle (1986) identified organizational commitment as a highly significant (p < .0005), negative correlate of turnover on the basis of adding z-values in the 16 samples they reviewed. Mathieu and Zajac (1990) combined data from 26 samples and reported a mean weighted correlation corrected for attenuation between organizational commitment and turnover of -.28. Stronger effects were reported on intention to search for job alternatives ((symbol omitted) = -.60, t = 5 samples) and intention to leave a job alternatives ( (symbol omitted)
= -.46, t = 36 samples), variables believed to mediate the commitment turnover relationship.
Despite the abundance of research on the relationship between commitment and turnover, important problems hamper understanding of the causal processes involved. First, considerable disagreement about the concept of organizational commitment remains. Commitment theorists remain divided about the dimensionality of attitudinal commitment (Mathieu & Zajac, 1990). Second, disagreement persists concerning the nature of the psychological withdrawal process that employees experience prior to quitting an organization. Mobley, Horner, and Hollingsworth's (1978) model has dominated research in this area (Williams & Hazer, 1986), but recent research has questioned the withdrawal process proposed in that model (Hom & Griffeth, 1991). Third, researchers have questioned the statistical appropriateness of using the path-analytic and structural equation modeling techniques relied on by most researchers to estimate commitment-turnover relationships (Huselid & Day, 1991).
The purpose of this study was to expand understanding of the commitment-turnover...