Figure 7-19 Structural specification for Model
The structural relationships in Model 5 are specified as illustrated in Figure 7 - 19
and the measurement model is initially specified (as per previous models) with all
of the measured (indicator) variables loading on their respective factors. This
initial specification is used, firstly for the model-based data imputation procedure
(to replace the missing data) and, secondly, to estimate and test the
measurement model in the usual way.
In Section 5.2, the EM construct was introduced as containing two sub-factors,
EM-I (Influence) and EM-C (Competencies). This conceptualisation reflected the
nature of the scale developed by Hancer and George (2003). The Competencies
sub-factor was recognised as being somewhat speculative owing to the
requirement to locate and include a third indicator variable from beyond Hancer
and George’s conceptual framework; nb - this third indicator was identified from
Lundberg et al. (2009).
Section 5.2 notes that if the measurement model confirms EM-I and EM-C as
distinct and valid constructs that covary significantly with the other constructs in
the measurement model, then EM-I and EM-C will be modelled in the structural
model as sub-factors of a second-level Empowerment construct.
Following an initial development of the measurement model for Model 5, it was
found that the EM-C sub-factor did not covary significantly with either the
Motivational Leadership (p = 0.152) construct or the Discretionary Service
Behaviour construct (p = 0.190). Accordingly, the EM-C sub-factor and its
associated indicator variables (EM6, EM7 and EM8) were removed from the
analysis. This modification is not entirely unexpected (EM-C was included