For the modifications in Model 1, and throughout the model development process,
indicator variable removal was undertaken with close regard to the substantive
implications for construct identities. For example, the item JP4 (Job Performance
construct) was retained for substantive reasons (its service-focused content) in
the face of its moderate factor loading (details in Section 7.2.8). Accordingly,
following these modifications, the identity of each construct remained essentially
unchanged (post-modification construct identities are described in detail in
Section 7.13).
Aside from confirming hypotheses 1 and 2, Model 1 is important in demonstrating
the validity of the combined approach of measuring job performance using (i)
self-assessment (the JP construct) and (ii) peer-assessment (the DSB construct).
Specifically, the similarity between the two parameter estimates ( = 0.415 for
ML→JP and = 0.396 for JP→DSB) provides some assurance that the self-
assessed Job Performance scores are not unrealistically inflated (see Van Dyne
and LePine, 1998 and Simons and Roberson, 2003).
Because the JP construct (the individual’s performance) measures motivational
leadership outcomes at the individual level, and DSB (colleagues’ performance) at
the group level, an indication of motivational leadership’s positive outcomes at
multiple levels is provided.
Model 2 introduced the Work Meaning construct as a direct outcome of
Motivational Leadership and a partial mediator of Motivational Leadership’s effect
on Job Performance.