Abstract
Several data sets collected on the Mobley, Horner, and Hollingsworth (1978) model of turnover were re-analyzed with path analytic techniques. Data analyses revealed support for three general hypotheses proposed by Mobley et al. These were: (1) that age has an indirect effect on turnover through job satisfaction, (2) that job satisfaction has an indirect effect on turnover through withdrawal cognitions, and (3) that intention to quit is the immediate precursor of turnover. Although the data supported these hypotheses, inconsistencies were apparent in the data for many specific linkages proposed by Mobley et al. Possible reasons for these inconsistencies include: (1) diversity of samples studied, (2) the unmeasured variables problem, and (3) construct measurement problems. The use of a causal modeling approach developed by Joreskog & Sorborm (1978) is proposed as a possible solution to some of these problems in future research.
Get full access to this article
View all access options for this article.
