Leader Member Exchanges: Distinction Between Two Factors

Abstract

The dimensionality of Graen's Leader Member Exchange concept was examined in a cross-sectional questionnaire study among 1147 Dutch nurses. The total sample was divided into two subgroups by means of stratified random selection (n1 = 574; n2 = 573) and analyses were performed in each subgroup separately. The factorial structure of a 12-item self-report Leader Member Exchange questionnaire was examined. Exploratory factor analysis on the data of subgroup 1 yielded two factors, which were labelled “consideration/coaching by the leader” and “filling an assistant position”. Confirmatory factor analyses with LISREL showed that a two-factor oblique model had the best fit in both subgroups. Equality of factor structures in both subgroups was clearly demonstrated. Differential validity of the two Leader Member Exchange factors was studied by correlating each factor with measures of job-related affect. In both subgroups, the consideration/coaching factor had a significantly stronger, positive relationship with job satisfaction and a significantly stronger, negative relationship with propensity to quit the job, whereas the assistant position factor had a significantly stronger, positive relationship with job-responsibility. Results concerning the relationships with effectiveness ratings were inconclusive. The results of this study indicate that the Leader Member Exchange concept is two-dimensional rather than unidimensional.