Discussion
The one-dimensional model yielded good fit indices, showed high factor
loadings, and excellent internal consistency. Moreover, bivariate correlations supported
convergent and discriminant validity of the PLCQ-OR. Although the correlations
with LMX-A and TL-SL are high, item content clearly indicates conceptual
differences. For instance, LMX-A includes items such as “I like my
supervisor very much as a person” or “My supervisor is a lot of fun to work
with,” and the TL-SL consists of items such as “Behaves in a manner which is
thoughtful of my personal needs”—neither the LMX nor the TL contain items
that explicitly reflect communicative behavior. Moreover, current leadership theories
cannot easily incorporate such items because no theoretical assumptions are
made about leadership communication. Therefore, these scales cannot be used
to analyze (and optimize) communicative aspects of leadership. Last, the positive
relationship with the criterion variable job satisfaction indicates concurrent
validity