Results and discussion
In the end, the model resulted in five working constructs: job satisfaction, perceived justice, perceived support (encompassing POS and PSS), intrinsic rewards, and extrinsic rewards. According to Coomber and Barriball, (2007), job satisfaction can be studied as a whole concept (global approach) or focussing on specific areas ( facet approach). This study uses the global approach and a multi-item construct (Saks, 2006) because of the many positions under analysis.
Table II describes the final constructs. Although PSS and POS were to be considered separately, the model showed a better fit when combining all the support-related items in one construct, even though all PSS items present higher mean values than the POS items. This concurs because POS influences the quality of the supervisor-subordinate relationship, and therefore the PSS (Eisenberger et al., 2002; Wayne et al., 1997). Regarding the items of the perceived justice and fairness construct, Table II shows that unbiased job decisions have the highest mean value, whereas the lowest mean value belongs to having all job decisions consistently applied to all affected employees. Among the intrinsic rewards, public employees seem to prefer the perception of respect from their colleagues, whereas having more challenging work assignments comes in last place. As for the extrinsic rewards, Table II also confirms that pay rises are less valued than other rewards and compensations.
