Result of Hypotheses Testing

4.3.1 Results on direct antecedent of job satisfaction, turnover intention

Regression analysis is also conducted to test the hypothesis and to define the direction of relations. In this section, regression analysis was used to test the effect of POS include fairness treatment perceived, reward and job condition perceived toward job satisfaction – hypotheses H1, H1a, H1b and job satisfaction toward turnover intention – hypothesis H2 that were proposed in the last parts.

The results obtained from the regression analysis are shown in tables below.

From Table 4.9, the first finding is an R² value of the model 1 to test hypothesis H1 was calculated as .588. This value shows the perceived organizational support explained 58.8%

variability of job satisfaction with  = .767 (p < .001). It can understand as research model show F value equal to 383.933 which is significant due to the value is less than .001. Hence hypothesis H1 “Perceived organizational support will positively affect job satisfaction” is supported.

Table 4.9 Regression analysis results of direct antecedent on Job Satisfaction, Turnover Intention

Testing Dependent variable Independent

variable Beta R-square F

H1 POS Job satisfaction .767*** .588 383.933***

H1a Perceived fairness Job satisfaction .081

.602 202.565***

H1b Perceived reward and job condition

Job satisfaction .723***

H2 Job satisfaction Turnover intention -.394*** .152 49.341***

*p<.05 **p<.01 ***p<.001

To know the effect of POS to job satisfaction is fully supported or only partial support the analysis relationship of 2- factors of POS which are perceived fairness treatment and perceived organizational reward and job condition were explored in the factor also using this phrase. The result was showed in Table 4.9 that perceived fairness support have no affecting job satisfaction ( = .81, p > .05) whereas, perceived organizational reward and job condition support show positive affect job satisfaction ( = .733, p < .001) and can explain 60.2% of the variance in job satisfaction (R² = 0. 602, F= 202.565). Thus the supplementary hypothesis H1b is accepted and H1a isn’t accepted so POS is partial effect to job satisfaction. From that finding, the researchers point out company should pay more attention to an important role of organizational reward and job condition in order to enhance job satisfaction.

Job satisfaction shows the significant negative effect to turnover intention from the finding in the same table above. R² value of the model 2 to test hypothesis H2 was calculated as .152 so job satisfaction may explain 15.2% variability of turnover intention with  = .767, F= 49.341, p < .001, it mean the higher level job satisfaction employee have, the less likely they are to leave. In other words, “Job satisfaction will be negatively related to turnover intention” – the hypothesis H2 is accepted.

In conclusion, hypotheses H1, H1b, H2 are all accepted except H1a.

4.3.2 Mediating regression analysis of Job satisfaction

According to Baron & Kenny (1986), Judd & Kenny (1981) mediation is supported when four conditions are met: (1) The relationship between the independent variable and the dependent variable must be significant. (2) The relationship between the independent variable and the mediator/intervening variables must be significant. (3) The relationship between the mediator variable and the dependent variable must be significant. (4) The relationship between the independent variable and the dependent variable must be weaker (partial mediation) or insignificant (full mediation) in the presence of the mediating variable. In

another hand, many recently researchers and typically Shrout & Bolger (2002) point that step 1 is not necessarily required in process of mediator testing so this study uses the guidelines of Shrout & Bolger (2002) with 3 conditions to test a mediating role of job satisfaction.

Hypothesis H3 has proposed that job satisfaction mediates the relationship between POS and turnover intention. Hence, in this testing, perceived fairness support, and perceived reward, job condition support are independent variables, job satisfaction is mediating variable and turnover intention is a dependent variable. The results are demonstrated in Table 4.10.

Table 4.10 Regression analysis results of Job satisfaction mediating Job satisfaction Turnover Intention reward and job condition support show the significant relationship with the mediator factor (= .723, p < .001) and perceived fairness support do not show the direct effect to job satisfaction (= .081, p > .05). Secondly, a result of testing Step 2 pointed that the mediator factor job satisfaction has a significant relationship with dependent variable – turnover intention ( = -.394, p < .001). And finally, Step 3 in this table show a significant mediator of job satisfaction on POS and turnover intention relationship with = - .183, p < 0.05, F=32.628, p<0.001 and R² = 0. 268.

In detail, perceived fairness treatment dimension show a significance in the direct path from POS to turnover intention (= - .412, p < .001) also significant but decreasing effect in the indirect path from it to turnover intention when job satisfaction was controlled ( = -.398, p

< .001) but it not adapt one condition of mediating testing which is a relationship with mediator factor so job satisfaction do not mediate the relationship between perceived fairness treatment and turnover intention.

Whereas perceived organizational reward and job condition has positive affecting to mediator factor and have a significant direct path to turnover intention ( = - .131, p < .05) but show a no significance in the indirect path to turnover intention when job satisfaction was controlled ( = .001, p > .05) thus it can conclude job satisfaction fully mediate the relationship between perceived reward and job condition to turnover intention – and job satisfaction only partially support the relationship between overall POS to turnover intention, in other words, hypothesis H3 is accepted.

4.3.3 Moderating regression analysis of Organizational reputation

In order to understand the role of organizational reputation on predicting turnover intention-behavior, in the case of this research that is “bad reputation” of Formosa Ha Tinh Steel, Vietnam, hypothesis H4 “Organizational reputation will moderate the relationship between job satisfaction and turnover intention” was proposed. This section will discuss it.

Table 4.11 Results of Organizational reputation moderating role Turnover Intention

In a causal relationship, if x is the predictor variable and y is an outcome variable, then z

is the moderator variable that affects the casual relationship of x and y if it found to be significant and can cause an amplifying (> 0, p < .05) or weaken (< 0, p < .05) effect between independent and dependent variable.

As shown in Table 4.11, the model testing with 3 steps. Firstly, same as hypothesis H2 - the interaction of job satisfaction and turnover intention was significant (= -.394, p < .001) and it still significant when organizational reputation occur in the regression because the result show effect of job satisfaction decrease (= -.368, p < .001) although organizational reputation wasn’t significance (= .111, p > .05) relate with intending turnover at step 2 in testing model. Step 3 show the moderator role of organizational reputation was significance on the relationship between job satisfaction and turnover intention (=.261, p< .001) Additionally, job satisfaction has a slightly descending affect to turnover (= -.370, p < .001) compare with step 2 and still decreasing affecting to turnover if consider organizational reputation as moderator in the regression compare with step 1 (Δ= -.024). Hence, organizational reputation can amplify the relationship between job satisfaction and turnover intention with the more increasing organizational reputation the stronger affecting of job satisfaction toward turnover intention behavior. In this case, the rising bad reputation of the company can cause to employee’ embarrassment, discomfort, decrease their self-esteem and satisfaction before, finally affect to their turnover intention behavior.

In sum, hypothesis H4 “Organizational reputation will moderate the relationship between job satisfaction and turnover intention” is accepted.

4.4 Comparing the demographic characteristic to POS and