Figure 3. Path Coefficient Inner Model
From the results of hypothesis testing in table 4 above, it is known that the regression coefficient of PS-Fit’s effect on LMX (β= 0.612; t= 10,631; p<0.010) indicates that PS-Fit has a positive effect on LMX. The results of testing the hypothesis show support for the first hypothesis, which says that PS-Fit has a positive effect on LMX. This means that the better the PS-Fit in the Islamic bank, the closer the leader and subordinates will be.
From the results of hypothesis testing in table 4 above, it is known that the regression coefficient of the effect of LMX on employee performance (β= 0.415; t= 5.296; p<0.010) shows that LMX has a positive effect on employee performance. The results of testing the hypothesis support the second hypothesis, which says that LMX has a positive effect on employee performance. This means that the better the LMX, the better the employee performance of Islamic banks.
From the results of hypothesis testing in table 4 above, it is known that the regression coefficient of the effect of LMX on work engagement is (β= 0.387; t= 5.062; p<0.010), which shows that LMX has a positive effect on work engagement. The results of testing the hypothesis show support for the third hypothesis, which says that LMX has a positive effect on work engagement. This means that the better the LMX, the higher the work engagement of Islamic banks.
From the results of hypothesis testing in table 4 above, it is known that the regression coefficient of the effect of PS-Fit on employee performance through LMX is (β= 0.254; t= 4.772; p<0.010), which indicates that LMX is able to mediate the effect of PS-Fit on employee performance. These results indicate the role of LMX as a partial mediation of the effect of PS-Fit on employee performance. This is because the influence of PS-Fit on employee performance is significant either directly or through LMX. The results of hypothesis testing support the fourth hypothesis, which explains that LMX mediates the effect of PS-Fit on employee performance.
From the results of hypothesis testing in table 4 above, it is known that the regression coefficient of the influence of PS-Fit on work engagement through LMX is (β= 0.237; t= 4.656; p<0.010), which indicates that LMX is able to mediate the effect of PS-Fit on work engagement. These results indicate the role of LMX as a partial mediation of the effect of PS-Fit on work engagement. This is because the influence of PS-Fit on work engagement is significant either directly or through LMX. The results of testing these hypotheses support the fifth hypothesis, which explains that LMX mediates the effect of PS-Fit on work engagement.