Despite the similarity of factor structures between ME and EPA, the new construct
was not identified as a measure of work meaning because of the inclusion of the
AOC2 indicator which relates to employees’ emotional attachment to their
organisation. The specific impact of this last point is that no existing
conceptualisations of work meaning that have been identified include any
component related to emotional attachment – indeed, emotional attachment is
specific to Affective Organisational Commitment.
In substantive terms, and following from the findings of Model 2, Model 3b tells us
that motivational leaders, in addition to enhancing work meaning for employees,
enhance a broader class of positive attitudes to work and that these positive
attitudes in turn give rise to enhanced job performance by employees.
7.7 Model
Model 4 sees the introduction of the Work Values construct to the modelling
process. Once again, the reformation of the Employee Attitude variables into the
new EPA construct requires a new hypothesis to be articulated. Accordingly, in
place of H 9 , H 10 and H 11 (which linked Work Values with the original ME, JS and
AOC constructs), H 23 now states that:
employees with more positive dispositions towards work in general will also
report higher levels of positive work-related attitudes
Broad theoretical support for H 23 comes from the values→attitudes→behaviour
(V→A→B) hierarchy described by Homer and Kahle (1988: 638-639).
The structural model is specified as illustrated in Figure 7 - 17 and tests H 23 (as
described above) and re-tests H 1 , H 2 and H 22 within the context of the latest
model configuration. Once again, the measurement model is initially specified
with all of the measured (indicator) variables loading onto their associated factor
(5 indicators for ML, 4 for JP and DSB, 14 for EPA and 7 on Work Values).
The first step is to specify the measurement model (with the indicators as
described above and with double-headed covariance arrows inter-connecting each
construct) and to use this configuration to replace the missing values using
AMOS’s model-based imputation method.
