values. The resulting distributions across the missing / less / more categories are
described in Table 7 - 40.
Variable Work characteristic categories
Contact
(Missing values n = 39 )
Less More
76 98
Training
(Missing values n = 48 )
Less More
107 58
Information
(Missing values n = 43 )
Less More
114 66
Table 7-40 Group distributions for work characteristic categories
Once again, the multi-group analyses for moderator effects were carried out on
both Model 2 and Model 5b.
For Model 2:
contact with supervisor - model fit was good for both the measurement and
structural multi-group models and no moderation effect was found;
adequacy of training – findings suggest no moderation effect but model fit is
less than satisfactory for the structural model; and
adequacy of information – findings do suggest a moderation effect but model
fit is not satisfactory for either the measurement or structural models.
For the more complex Model 5b, model fit was poor for measurement and
structural models for each of the three groupings. As with Model 2, adequacy of
information was signalled as producing a moderation effect but the poor model fit
statistics prevent anything more than a cautious interpretation of this effect.
In summary, the only finding that can be confidently concluded from the analyses
reported above is that age does not moderate any of the relationships in Model 2.
The lack of adequacy in achieving model fit for these multi-group models is likely
a reflection of the smaller sample sizes in comparison with the demographic-
based multi-group models examined in the previous section.
7.15.3 Work values
Section 2.3.2 introduced the humanistic tradition in leadership studies. This
strand of research recognised employees’ personal objectives as contributors to
organisational and individuals’ outcomes. The notion that organisations and