method. All of the model parameters were found to be robust to the non-normal
multivariate distribution as described in Appendix IV.
Model 4 mediation and total effects
A constrained model with no direct effect (ML→JP) was estimated (SEM 4:2 1 ).
The chi square difference (Δχ^2 ) between models SEM 4:2 (χ^2 = 120.67) and SEM
4:2 1 (χ^2 = 111.355) was calculated at 9.315. With 1 degree of freedom this
yields a p value of 0.0023. Accordingly, (because p <0.01) full mediation of the
ML→JP path by EPA is not supported.
To determine whether or not Employee Positive Attitudes (EPA) has a partial
mediation effect on the ML→JP relationship it is necessary to estimate a third
model (SEM 4:2 2 ) in which there is no indirect effect (EPA→JP) on Job
Performance (JP).
The structural coefficient values for the ML→JP path are now compared between
model SEM 4:2 2 and the hypothesised model SEM 4:2. In the constrained model
(SEM 4:2 2 ) the value (effect size) for the ML→JP path is 0.424 while in the
hypothesised model (SEM 4:2) this value is 0.259. When comparing the ML→JP
path between the constrained model (SEM 4:2 2 ) with the hypothesised model,
because ML→JP is statistically significant in both models but the parameter
estimate is reduced in size in the hypothesised model (0.424 in SEM 4:2 2 and
0.259 in SEM 4:2) we can conclude that partial mediation is supported (see Hair
et al. 2006 : 867).
Table 7 - 24 describes the total (combined direct and indirect) effects for Model 4
(SEM 4:2). In comparison to Model 3b, the total effect of ML on EPA has dropped
considerably (from 0.574 to 0.446), the total effect of ML on JP is slightly down
(from 0.415 to 0.380) and ML→DSB is much the same (0.403 versus 0.401 in
Model 3). Also remaining at a similar level is the EPA→JP path (0.273 versus
0.267 in Model 3).
In Model 4, there is a second mediation effect to be tested – that is, the
hypothesised full mediation of the WV→JP path by EPA. To confirm the full
mediation effect of EPA - following Hair et al.’s guidelines (2006: top of p. 868) –
a direct path between WV and JP is estimated. Because this path is not
significant (p = 0.139) the full mediation of the WV→JP path by EPA is supported.