so the regression model fulfills the normality assumption and regression model that is used to predict the service quality and satisfaction based on
customer loyalty.
G. Multiple Linier Regression
1. Determination Coefficient R
2
Table 4.36 Model
R R
Square Adjusted
R Square
1 0.647
0.418 0.393
a. Predictors: constant, Cust_Sat, Serv_Q b. Dependent Variable: Cust_loyal
From the model summary above, we can see that determination of coefficient R Square around 0.418 means that the services quality and
satisfaction has influence to customer loyalty at 41.8 while the rest is 58.2 influenced by another variable that is unknown and not included in
this regression analysis.
2. Similarity of Multiple Linier Regression
Table 4. 37 Understandardized
Coefficient Model
B Std.Error
1 constant 0.926
3.003 serv_Q
0.027 0.04
Cust_Sat 0.459
0.116
The result of the coefficient regression shows that constantan coefficient value is 0.926 and with the T test is 0.308. The significant
value is 0.759. The slope coefficient of services quality is 0.027 with the T test of 0.672 and significant value of 0.505.and coefficient slope of
customer satisfaction of 0.459 with T test value of 3.949 and significant value of 0.000.
The result is the regression similarity: Y = 0.926 + 0.027 X
1
+ 0.459 X
2
Where: Y = customer loyalty
X
1
= services quality X
2
= satisfaction From the similarity means that the price of 0.926 is the constant
value shows that the factors influence the services quality and satisfaction, the value of customer loyalty around 0.926.
The service quality variable X
1
have positive influence to customer loyalty, with coefficient regression about 0.027, means there is
effect on increasing customer loyalty about 0.027. The satisfaction variable X
2
have the positive influence to customer loyalty, with the coefficient regression around 0.459, means that
there is the effect of increasing the customer loyalty around 0.459.
3. F Test
Table 4. 38
ANOVA
b
46.918 2
23.459 16.895
.000
a
65.262 47
1.389 112.180
49 Regression
Residual Total
Model 1
Sum of Squares
df Mean Square
F Sig.
Predictors: Constant, Cust_Sat, Serv_Q a.
Dependent Variable: Cust_Loyal b.
Based on the table 4.38 on the above, it is shown that the multiple coefficient regression result by F test is 16.895 while the significant level
shows 0.000 0.05 and F table is 3.23 with the independent degree of 47, so the conclusion that F test
F table
,
means that significant. It means Ho is rejected and Ha is accepted. It can be concluded that there is significant
influence between services quality and satisfaction to customer loyalty in BNI.
4. T Test
Table 4.39 Coefficients
Model t
Sig.
1 Constant .308
.759 Serv_Q
.672 .505
Cust_Sat 3,949
000
a. Dependent Variable: Cust_Loyal
Based on the coefficient table on the above, to know the influence of each dependent variable, such as:
1. The variable X
1
service quality T test 0.672 T table 1.684 and significant value is 0.505 0.05. That means Ha is rejected, so that the
services quality variable does not influence to variable Y customer loyalty.
2. To variable X
2
satisfaction T test 3.949 T table 1.684 and significant is 0.000 0.05. Means Ha is accepted, so the satisfaction
variable is influence by significantly to variable Y customer loyalty.