*3.3. Discriminant Analysis*

We present the summary of the results of the discriminant analysis in Table 2. The results show that employees with the unsatisfied–unhealthy pattern (Cluster 1) have systematically higher means on role ambiguity and lower means on job importance, compared to employees with the satisfied-healthy wellbeing pattern (Cluster 4) and the rest of the patterns. Therefore, we partially confirmed hypothesis 1. Thus, role ambiguity and job importance strongly differentiated between the unsatisfied-unhealthy and satisfied-healthy patterns, but we failed to confirm role conflict. Contrary to our expectations, role conflict mainly characterized satisfied-unhealthy employees. Discriminant results also show that employees with the satisfied-unhealthy pattern (Cluster 3), systematically had significantly higher means on role overload (and role conflict), compared to employees with the satisfied-healthy pattern (Cluster 4). Therefore, we confirmed hypothesis 2, which stated that role overload characterized employees with the satisfied-unhealthy pattern. Finally, employees with the unsatisfied-healthy pattern (Cluster 2) have significantly higher means on overqualification (in comparison with the rest of the patterns) and lower means on role conflict, compared to the satisfied-unhealthy pattern (Cluster 3). Therefore, we also confirmed our hypothesis 3; thus, employees with the unsatisfied-healthy pattern perceived themselves as more overqualified for the job/position compared to their satisfied-unhealthy counterparts and the rest of the patterns.


**Figure 2.** Four-cluster solution: standardized cluster means.


#### *IJERPH* **2019** , *16*, 25

**Table 2.**

Discriminant

 analysis of the four patterns of relations between job satisfaction

 and mental health with job importance,

 role ambiguity, role conflict, role

Together, the results of the discriminant analyses suggest that job importance, role ambiguity, role conflict, role overload, and overqualification help to differentiate among the four patterns of relations between job satisfaction and mental health. When comparing these variables, job importance and role ambiguity were better at differentiating employees with the unsatisfied-unhealthy pattern from those with the satisfied-healthy pattern. Overqualification and role conflict were better at differentiating between employees with the unsatisfied-healthy and satisfied-unhealthy patterns.
